The magnetic response of most natural materials, characterized by magnetic permeability, is generally weak. Particularly, in the optical range, the weakness of magnetic effects is directly related to the asymmetry between electric and magnetic charges. Harnessing artificial magnetism started with a pursuit of metamaterial design exhibiting magnetic properties. The first demonstration of artificial magnetism was given by a plasmonic nanostructure called split-ring resonators. Engineered circulating currents form magnetic plasmons, acting as the source of artificial magnetism in response to external electromagnetic excitation. In the past two decades, magnetic plasmons supported by plasmonic nanostructures have become an active topic of study. This Perspective reviews the latest studies on magnetic plasmons in plasmonic nanostructures. A comprehensive summary of various plasmonic nanostructures supporting magnetic plasmons, including split-ring resonators, metal–insulator–metal structures, metallic deep groove arrays, and plasmonic nanoclusters, is presented. Fundamental studies and applications based on magnetic plasmons are discussed. The formidable challenges and the prospects of the future study directions on developing magnetic plasmonic nanostructures are proposed.

Any electromagnetic (EM) waves consist of oscillating electric and magnetic fields. These two parts are inherently connected together with one never occurring without the other. However, when interacting with conventional materials, these two components behave very differently: the electric field drives the charges in the atoms more efficiently, leading to a strong electric response of the materials. Whereas the magnetic counterpart interacts with the material very weakly because the driving force on the charges exerted by the magnetic field is four orders of magnitude smaller than that induced by the electric field.1 This is exactly the reason for the famous argument of Landau and Lifshitz: “there is certainly no meaning in using the magnetic susceptibility from optical frequencies onward, and in the discussion of such phenomena, we must put μ=1.”2 

In the field of plasmonics, such a feature of EM radiation still holds and has been fully employed to tailor the optical response of plasmonic nanostructures: the electric component of the EM field couples with the free charges on the surface of the metallic nanostructure, leading to surface-bound electromagnetic waves on metals known as localized surface plasmons (LSPs) or propagating surface plasmons polaritons (SPPs) [left panel in Fig. 1(a)].3,4 The plasmonic resonances have a unique feature of concentrating light in sub-wavelength scale. The tremendous boosting of the electric field at the nanoscale has found many promising applications such as biosensing,5 photo-detecting,6 light-harvesting,7,8 information processing,9–11 and non-linear light generation.5 

FIG. 1.

(a) Sketches of the excitations of SPP (top), LSP (middle), and MP (bottom). All these excitations are associated with a collective motion of surface charges under the illumination of light. (b) Diagram of MP-based plasmonic nanostructures that are used for both fundamental studies and their applications in various research fields.

FIG. 1.

(a) Sketches of the excitations of SPP (top), LSP (middle), and MP (bottom). All these excitations are associated with a collective motion of surface charges under the illumination of light. (b) Diagram of MP-based plasmonic nanostructures that are used for both fundamental studies and their applications in various research fields.

Close modal

In contrast, the achievement of a strong magnetic effect in plasmonic materials is a challenge due to the much weaker response to the magnetic component of light, which results from the small value of the magnetic permeability for most natural materials, particularly, at optical frequencies. The weak magnetic response in natural materials has motivated tremendous studies on “non-natural” structures and systems that may exhibit artificial magnetic properties. A typical way of generating artificial magnetism is to obtain engineered displacement currents and conduction currents in metamaterials, i.e., man-made composite materials constructed from subwavelength non-magnetic building blocks patterned into an effective material. Metamaterials with rationally designed properties can allow the magnetic component of the EM wave to interact with the meta-atoms, enabling strong and unique magnetic response with enhanced magnetic dipole moments even though the constituting elements are not endowed with microscopic magnetization.12,13

A pioneering work of demonstrating artificial magnetism in metamaterials was reported in 1999 by Pendry et al., who have proposed to exploit the inductive response from structured metallic non-magnetic materials, referred to as split-ring resonators (SRRs), to obtain high-frequency magnetism.14 SRRs possess a feature size smaller than the wavelength of radiation. Although there are no free magnetic poles in such systems, the excitation of displacement currents in the SRR can lead to induction of a magnetic dipole moment in response to the magnetic component of the incident EM field, as shown in Fig. 1(a)(bottom left). Similar to the electric plasmons excited in nanosized particles, such magnetic response also occurs with the generation of collective motion of charges on the metal surface. In this sense, an effective media made of SRRs can exhibit effective permeability that converges at a resonance defined as “magnetic plasma frequency ωMP.” This is the initial concept of “magnetic plasmons (MPs).” At the early stage of research in artificial magnetism, MP resonances excited at the long wavelength range were mainly demonstrated in SRR-based metamaterials with different configurations for realizing “left-handed” materials or negative-index materials (NIMs) at GHz frequency,15 or microwave cloaking devices.16 MP resonances can still be excited at a shorter wavelength at the infrared range by directly shrinking down the SRR size. But, it generally fails in the visible range due to increased radiation losses and fabrication issues.17 Metallic nanorod pairs, which were then transformed into metal–insulator–metal (MIM) structures such as “fishnet,” were, therefore, proposed to overcome these difficulties. The MP excitation in those metamaterials can be regarded as a resonant inductor–capacitor (LC) circuit with a current loop operating at an optical range, leading to optical magnetism at LC frequency, which is particularly useful for realizing optical NIMs.18 However, as the structural configurations are fixed the MP property can hardly change, which is not applicable for controllable magnetism.

Another category of magnetic nanostructures is plasmonic nanoclusters with the multiparticle arrangement of plasmonic nanoparticles (NPs), also called metamolecules or plasmonic oligomers. Their MP properties are determined by the mutual interaction between the constituting sub-units, which provide flexible control of the artificial magnetism. As the plasmonic NPs are closely packed together and arranged in circular loops, MP can be excited as a result of resonant displacement contribution. The rational design of nanoclusters can help us to achieve magnetic resonances at visible frequencies with flexible control both in their resonance positions and their spectral profile, even with lower intrinsic losses.19 Moreover, the MP resonance possesses higher sensitivity to the local dielectric environment and enhanced magnetic fields at the center of the nanostructure assembly.20 All these advantages make the magnetic nanoclusters promising platforms for both fundamental research and practical application in many fields, such as light-harvesting,21 MP propagation,22 and biodetection.23 

So far, different approaches have been followed to achieve MP generation and advanced manipulation on the magnetic response of plasmonic nanostructures at the nanoscale level, as summarized in Fig. 1(b). Artificial magnetism is required to be excited in nanostructures with simpler configurations such as 1D deep groove arrays.24 Mutual interactions of MPs, or coupling of MP with other optical excitations (such as quantum emitters and optical cavities), has also drawn much attention due to their potential use for functional nanophotonic devices or active plasmonic devices.24,25 It is worth noting that all-dielectric nanoparticles, as an alternative strategy to generate artificial magnetism, have also been investigated due to their intrinsic low-loss feature.26–29 A short review on optical magnetism supported by all-dielectric nanostructures can be found in Ref. 30. Considering the length limitation of this Perspective, we only focus on plasmonic nanostructures.

This Perspective covers most typical categories of MP-based plasmonic nanostructures with artificial magnetism and fascinating applications. Section I contains a brief introduction to the field of plasmonics and highlights important findings associated with MP generation and applications. Section II discusses the effects and applications of MPs in different plasmonic nanostructures. We begin in Sec. II A with an overview of the basic working principles of MP generation in SRRs and review its development in negative-index materials and the current research progress in modern metamaterials. In Secs. II B and II C, we review the MP excitations in metal–insulator–metal systems, with particular emphasis on the “fishnet” nanostructures and nanoparicle-on-mirror systems. In Sec. II D, the MP generation in deep groove arrays is discussed. Their applications in the field of light collection and sensing are demonstrated. Next, in Sec. II E, we introduce the mechanism of MP generation by nanoclusters and their powerful ability to manipulate the tunable optical properties associated with mutual particle coupling. At the end of this section, we dedicate Sec. II F to MP-based light–matter interactions, especially the exciting phenomena arising from the coupling of MP-producing nanostructures to two-dimensional materials. Finally, we present our outlook on the field and summarize recent advances that pave the way for practical magneto-optoelectronic devices.

One of the earliest nanostructures used to generate MP is the split-ring resonators (SRRs) composed of non-magnetic metals proposed by Pendry et al. in 1999.14 In this pioneering work, Pendry et al. demonstrated for the first time the concept of SRRs as an effective way of generating strong magnetic response for achieving negative permittivity at the gigahertz range of frequencies. Following this great idea, researchers have demonstrated a variety of nanostructures containing SRRs with different configurations to obtain a desired magnetic response for various applications. Here, we will first give a short discussion on the basic working principle of MP generation in SRRs. Then, we will give a short review of their early stage development in realizing negative permittivity for refractive index and left-handed materials. We will finally focus on the recent advances in the study of MP resonances in SRR-based metamaterials with novel functionalities.

The initial version of the SRR structure proposed by Pendry and his co-workers consists of two concentric SRRs with subwavelength dimensions, as depicted in Fig. 2(a) (left). These two SRRs are placed in opposite directions, which can be viewed as an equivalent inductor–capacitor (LC) circuit consisting of inductive and capacitive elements. The inductances are generated by the two metallic rings. The capacitors are formed by the two slits and the gap between the two rings. This circuit leads to an LC oscillator of eigenfrequency of MP resonance with negative permeability. The working principle for realizing negative permeability is as follows: the incident light provides a magnetic field oriented perpendicular to the plane of the rings. The magnetic field, according to Lenz’s law, induces an opposing magnetic field in the metallic rings. This gives rise to a diamagnetic response and hence to a negative real part of the permeability.31 

FIG. 2.

(a) Schematic diagram of two concentric SRRs with circle shapes (left) and square shapes (middle).15 Reprinted with permission from Shelby et al., Science 292, 77–79 (2001). Copyright 2001 The American Association for the Advancement of Science. SEM picture of the metasurface composed of two concentric SRRs and their units (right).32 Reprinted with permission from Yen et al., Science 303, 1494–1496 (2004). Copyright 2004 The American Association for the Advancement of Science. (b) Schematic diagram of single SRR (top left), SEM pictures of the metasurface composed of single SRRs (top middle), and the schematic diagram of the LC circuit of single SRR (top right).33 Reprinted with permission from Linden et al., Science 306, 1351–1353 (2004). Copyright 2004 The American Association for the Advancement of Science. SEM picture of SRR (bottom left), and transmission spectra of single SRRs at incident angles of 0° (black), 30° (red), 60° (green), and 90° (blue) (bottom right).34 Reprinted with permission from Enkrich et al., Phys. Rev. Lett. 95, 203901 (2005). Copyright 2005 American Physical Society. (c) Schematic diagram of the quasi-planar plasmonic metamaterial demonstrated by combining dumbbell aperture with VSRR, and (d) its unit.35 Reprinted with permission from Wu et al., ACS Nano 12, 1920–1927 (2018). Copyright 2018 American Chemical Society. (e) The reflection (green), transmission (red), and absorption (blue) spectra of VSRRs under the excitation of the normal incident light with its electric field polarized parallel to the VSRRs, (f) and its normalized electric field in the x–z plane.36 Reprinted with permission from Tsai et al., Adv. Mater. 31, 1806479 (2018) Copyright 2018 John Wiley and Sons. (g) Schematic diagram (g) and the top view SEM picture (h) of Au SRR structure on top of an MO thin film Ce-doped YIG (Ce:YIG).37 Reprinted with permission from Yang et al., Nat. Commun. 13, 1719 (2022). Copyright 2022 Springer Nature.

FIG. 2.

(a) Schematic diagram of two concentric SRRs with circle shapes (left) and square shapes (middle).15 Reprinted with permission from Shelby et al., Science 292, 77–79 (2001). Copyright 2001 The American Association for the Advancement of Science. SEM picture of the metasurface composed of two concentric SRRs and their units (right).32 Reprinted with permission from Yen et al., Science 303, 1494–1496 (2004). Copyright 2004 The American Association for the Advancement of Science. (b) Schematic diagram of single SRR (top left), SEM pictures of the metasurface composed of single SRRs (top middle), and the schematic diagram of the LC circuit of single SRR (top right).33 Reprinted with permission from Linden et al., Science 306, 1351–1353 (2004). Copyright 2004 The American Association for the Advancement of Science. SEM picture of SRR (bottom left), and transmission spectra of single SRRs at incident angles of 0° (black), 30° (red), 60° (green), and 90° (blue) (bottom right).34 Reprinted with permission from Enkrich et al., Phys. Rev. Lett. 95, 203901 (2005). Copyright 2005 American Physical Society. (c) Schematic diagram of the quasi-planar plasmonic metamaterial demonstrated by combining dumbbell aperture with VSRR, and (d) its unit.35 Reprinted with permission from Wu et al., ACS Nano 12, 1920–1927 (2018). Copyright 2018 American Chemical Society. (e) The reflection (green), transmission (red), and absorption (blue) spectra of VSRRs under the excitation of the normal incident light with its electric field polarized parallel to the VSRRs, (f) and its normalized electric field in the x–z plane.36 Reprinted with permission from Tsai et al., Adv. Mater. 31, 1806479 (2018) Copyright 2018 John Wiley and Sons. (g) Schematic diagram (g) and the top view SEM picture (h) of Au SRR structure on top of an MO thin film Ce-doped YIG (Ce:YIG).37 Reprinted with permission from Yang et al., Nat. Commun. 13, 1719 (2022). Copyright 2022 Springer Nature.

Close modal

The powerful capability of SRRs in tuning magnetic response in the gigahertz range was experimentally verified by Shelby et al. in 2001.15 The SRRs used in that work are in similar ring-like configuration as proposed in Pendry’s work, but with square shapes, as shown in Fig. 2(a) (middle panel). In such a double SRR structure, the generation of negative permeability originates from the excitation of MP induced by the circulating current in the square SRRs with the millimeter scale. Scaling this design down in size leads to similar magnetic responses at higher frequencies. MP resonance was excited by scaling the size of SRRs down to the micrometer scale (right panel), leading to negative permeability at 1 THz.32 To further achieve negative permeability at even higher frequencies, single SRRs [top panel in Fig. 2(b)] were employed with feature size scaled down to the nanometer. An array of single nonmagnetic metallic split rings generates a strong magnetic resonance, which arises from the LC circuit resonance at 100-terahertz frequency.33 In this sense, the equivalent LC circuit is much simpler (see top right panel). As the magnetic field vector of the incident light has a component normal to the plane of the ring, the circulating current, analogous to an atomic orbital current, induces a magnetic field that counteracts the driving magnetic field, thus leading to a negative permeability. Magnetic response at even higher frequencies was reported by employing a much simplified structure with a minimum feature size down to 50 nm [bottom left panel in Fig. 2(b)]. This so-called U-shaped structure is transformed from the traditional SRR by eliminating the tiny upper arms of the SRR. Arrays of such gold SRRs enable an MP resonance at 200 THz frequency [1.5 μm wavelength, bottom right panel in Fig. 2(b)], and even at higher frequency with the excitation of high-order MP around 370 THz (800 nm wavelength), almost reaching the visible range. Although the magnetic resonance frequency scales reciprocally with the structural size. At frequencies reaching the near-infrared and visible range, however, this linear scaling breaks down, i.e., one cannot achieve strong magnetic response by simply pushing the elementary SRR structures into an even smaller size.17 Magentic resonance operating at optical wavelength can be realized by designing other types of metamaterials, which will be discussed in later section.

In addition to the above-mentioned SRR-based structures, in which SRRs are arranged in 2D/3D arrays for realizing negative permeability or negative refractive index, SRRs can also be arranged in different ways to manipulate magnetic response for other functionalities and applications such as active optical metamaterials with tunable radiation loss at MP resonances and light manipulation, as well as MP-based sensing.38–47 The basic idea of such SRR-based nanostructures is arranging SRR units in an appropriate way to enable hybridization of different MP modes or interaction of MP with other optical excitations.39–43,48–51 For example, the excitation of MP resonances and their interactions can be greatly altered as the connection configuration or splitting part direction of SRRs in each unit changes. That is because each unit is no longer a SRR, but a resonator group formed by SRRs, which leads to the modification of the equivalent LC circuit with dramatically tailored magnetic response.32,46 One of the promising types of SRR-based nanostructures is metamaterials containing vertical SRRs (VSRRs).25,52–55 Compared to traditional SRRs fabricated on 2D planar subwavelength structures,47,56 3D vertically-aligned SRR metamaterials exhibit stronger field enhancement at the slit gaps due to the simultaneous excitation of both electric and magnetic dipoles with more adjustable and flexible control.57,58 All these advantages can be of great use for versatile applications such as plasmonic sensing,53,59 beam shaping60 and nonlinear effect in optical metamaterials.36 VSRRs have been successfully applied in constructing optical anapole metamaterial.35 A quasi-planar plasmonic metamaterial was demonstrated by combining dumbbell aperture with VSRR [Fig. 2(c)], in which transverse toroidal moment and resonant anapole behavior are present in the optical spectrum upon normal illumination of electromagnetic wave. VSRRs generate a strong magnetic dipolar response under normal illumination with the incident electric field parallel to the opening gap. The predominantly toroidal excitation is achieved by controlling the interaction between the magnetic dipole induced in VSRR and the two magnetic dipoles generated from the counter-rotating currents in the dumbbell aperture [Fig. 2(d)].

Recently, VSRRs have also been used for second harmonic light manipulation.36,61,62 Although the large field confinement of MP resonances in SRRs can provide a significant second harmonic generation (SHG) enhancement, such traditional planar SRRs suffer from inevitable leakage of electromagnetic energy into the underlying substrate, which greatly reduces the exposure of enhanced plasmonic fields at the structural interface. To solve this problem, Tsai et al. has recently demonstrated a 3D metastructure to enable stronger interactions, better electromagnetic field confinement, and less leakage into the substrate for great SHG enhancement.36 As shown in Fig. 2(e) (inset), when VSRRs are excited by normal incident light with its electric field polarized parallel to the VSRR while the magnetic field is perpendicular to the opening gap of VSRR, both the incident electric and magnetic fields offer required projection component for exciting the MP resonance of the VSRRs. The interplay between the electric dipole and magnetic dipole generates stronger localized fields within the opening gaps [Fig. 2(f)].

Another unique magnetic property called non-zero off-diagonal gyromagnetic permeability was recently realized in a magneto-plasmonic metamaterial.37 The basic idea of the proposed structure is the hybridization of SRR-based metamaterials with gyrotropic materials: a classical Au split-ring resonator (SRR) structure on top of an MO thin film Ce-doped YIG (Ce:YIG), as shown in Fig. 2(g). The microscopic mechanism of the emergent gyromagnetic property is attributed to the local electric/magnetic field direction modulated by the SRR arrays [Fig. 2(h)], underscoring a general strategy of introducing gyromagnetic properties in optical frequency metamaterials.

As mentioned above, the SRR-based metamaterials can support strong MP resonances at lower frequencies in the THz range for the realization of a negative index. However, the extension of such metamaterials to higher frequencies (particularly, the visible range) is restricted by the magnetic response saturation and the involvement of complicated and often difficult fabrication for the nanoscale metallic structures. An alternative way of accessing strong MP resonances at optical frequencies was the employment of metallic nanorod pairs, which were then transformed into metal–insulator–metal (MIM) structures with different configurations.63–66 One of the most popular MIM structures is the so-called “fishnet” structure, which has been widely used for RIMs at optical frequencies. As an important type of metamaterial that supports strong MP responses, the “fishnet” structure has been explored intensively over decades.64–70 Strong magnetic and electric response with flexible tunablility makes the “fishnet” structure a promising platform for the construction of functional metamaterials.

The study of “fishnet” structures initially started with an array of metal nanorod pairs.63 Early theoretical studies have predicted that a pair of metal nanorods can support a large paramagnetic/diamagnetic response and even a negative refractive index in the optical frequency range.76–79 Using the NIMs can realize the effect on the hidden body of the object,80 achieve the subwavelength resolution imaging beyond the diffraction limit, and so on.81,82 The first experimental demonstrations of negative refractive index in the optical range were reported by using pairs of metal rods and their transformed configurations, i.e., pairs of dielectric voids in metal.63,71 The metal rods can be considered inductors, and the gaps at the ends form two capacitors. The MP excitation is actually a resonant LC circuit with a current loop operating at optical frequencies.76,78,83 The mechanism of MP generation is as follows: In a nanorod pair [left in Fig. 3(a)] under the illumination of an electromagnetic plane wave, parallel currents in both rods can be induced by the oscillating electric field parallel to both rods. The magnetic field oriented perpendicular to the plane of the rods leads to antiparallel currents in the two rods, which in turn results in the strong MP resonance of the system. The magnetic response can be dia- or paramagnetic depending on whether the wavelength of the incoming magnetic field is shorter or longer than the MP resonance of the coupled rods. Using an electron-beam-fabricated sample of paired rods [right in Fig. 3(a)], Zhang et al. successfully realized a negative real part of refractive index n=0.3±0.1 in the optical range at telecommunication wavelength λ=1.5μm. However, this nanorod structure can only enable a partial negative refractive index due to the significant contribution from the imaginary part of the magnetic permeability, which only gives a low figure of merit defined as F=|n|/n with n being the imaginary part of the refractive index n.

FIG. 3.

(a) Schematic diagram of an array of nanorod pairs (left) and the field-emission SEM images of a fabricated array (right).63 Reprinted with permission from Shalaev et al., Opt. Lett. 30, 3356–3358 (2006). Copyright 2006 The Optical Society. (b) Schematic of a multilayer MIM structure consisting of an Al2O3 dielectric layer between two Au films perforated with a square array of holes (left) and SEM picture of the fabricated structure (right).71 Dark regions and hatched regions in the left panel represent the active electric and magnetic regions. Reprinted with permission from Zhang et al., Phys. Rev. Lett. 95, 137404 (2005). Copyright 2005 American Physical Society. (c) Schematic of a multilayer fishnet structure (left), and the corresponding side-view SEM image of the structure (right).72 Reprinted with permission from Zhang et al., Nature 455, 376 (2008). (d) Schematic of a fishnet metamaterial consisting of three unit cells in the propagation direction (left), and top-view SEM image of the fabricated fishnet structure (right).73 Reprinted with permission from Garcia et al., Phys. Rev. Lett. 106, 067402 (2011). Copyright 2011 American Physical Society. (e) A Fishnet structure with alumina spacer (left) and tilt-view SEM image of the structure after coating with Rh800-epoxy with a part of the top layer of silver removed by focused ion-beam milling (right).74 Reprinted with permission from Xiao et al., Nature 466, 735-U6 (2010). Copyright 2010 Springer Nature. (f) Geometry of a fishnet metamaterial (right) and measured transmission (black curve), reflection (blue curve), and absorption (red curve) spectra of the fishnet for horizontal polarization of the incident light (right).75 Reprinted with permission from Wang et al., ACS Photonics 3, 1494–1499 (2016). Copyright 2016 American Chemical Society.

FIG. 3.

(a) Schematic diagram of an array of nanorod pairs (left) and the field-emission SEM images of a fabricated array (right).63 Reprinted with permission from Shalaev et al., Opt. Lett. 30, 3356–3358 (2006). Copyright 2006 The Optical Society. (b) Schematic of a multilayer MIM structure consisting of an Al2O3 dielectric layer between two Au films perforated with a square array of holes (left) and SEM picture of the fabricated structure (right).71 Dark regions and hatched regions in the left panel represent the active electric and magnetic regions. Reprinted with permission from Zhang et al., Phys. Rev. Lett. 95, 137404 (2005). Copyright 2005 American Physical Society. (c) Schematic of a multilayer fishnet structure (left), and the corresponding side-view SEM image of the structure (right).72 Reprinted with permission from Zhang et al., Nature 455, 376 (2008). (d) Schematic of a fishnet metamaterial consisting of three unit cells in the propagation direction (left), and top-view SEM image of the fabricated fishnet structure (right).73 Reprinted with permission from Garcia et al., Phys. Rev. Lett. 106, 067402 (2011). Copyright 2011 American Physical Society. (e) A Fishnet structure with alumina spacer (left) and tilt-view SEM image of the structure after coating with Rh800-epoxy with a part of the top layer of silver removed by focused ion-beam milling (right).74 Reprinted with permission from Xiao et al., Nature 466, 735-U6 (2010). Copyright 2010 Springer Nature. (f) Geometry of a fishnet metamaterial (right) and measured transmission (black curve), reflection (blue curve), and absorption (red curve) spectra of the fishnet for horizontal polarization of the incident light (right).75 Reprinted with permission from Wang et al., ACS Photonics 3, 1494–1499 (2016). Copyright 2016 American Chemical Society.

Close modal

An alternative way of reaching NIM is the MIM design as the inverse configuration of the nanorod structure. A typical example is a pair of voids (nanoellipses) separated by a dielectric, as shown in Fig. 3(b) (left). The structure can be divided into two functional parts. The hatched regions mainly interact with the incident magnetic field. The top and bottom metal films generate a loop (inductor), which is then interrupted by the hole. As a result, a capacitor between the two films is also formed, which finally causes a strong MP resonance. Using this approach with fabricated nanovoids [right in Fig. 3(b)], Zhang et al. have obtained a negative refractive index at the wavelength λ=2μm with an improved figure of merit. With elliptically shaped voids etched in the two metal films, the figure of merit up to 2 was obtained.66 Later on these MIM configurations evolved to the “fishnet” structure to further improve the performance of NIMs. By replacing the circular or elliptical voids with rectangular counterpart, fishnet structures for NIMs were theoretically proposed and experimentally verified.64,65,68 The unit fishnet layer is functional with an antisymmetric current oscillation excited by the incident light field in the two metal layers of broad stripes. This causes part of a ring current, leading to a local magnetic dipole moment and thus to negative magnetic permeability above the magnetic resonance frequency. At the same time, because the pinstriped metal is parallel to the polarization direction of the excitation source, the oscillating frequency of the electron itself is lower than the effective plasma frequency, which results in the creation of a negative dielectric constant. The combination of these two aspects results in a negative index of refraction with low loss.16,64,65,68 Stacking the unit MIM fishnet unit vertically by many times can directly lead to a 3D-form fishnet structure.73 As shown in Fig. 3(d), Zhang et al. fabricated a 3D cascaded fishnet structure with a negative index existing over a broad spectral range. The structure [right in Fig. 3(d)] consists of 21 alternating layers of 30 nm silver (Ag) and 50nm magnesium fluoride (MgF2). This cascading causes a strong magneto-inductive coupling between neighboring functional layers. The tight coupling between adjacent LC resonators through mutual inductance leads to a broadband negative index of refraction with low loss, which originates from the destructive interference of the antisymmetric currents across the metal film and dramatically canceling out the current flow in the center of the film.84 

One step further to low-loss broadband NIM was realized at an optical wavelength (620806nm) with a high figure of merit up to 3.34. In contrast to other works that used first-order magnetic resonances, the basic idea is to utilize a second-order MP resonance of a fishnet structure, as shown in Fig. 3(c). With fewer layers (seven layers) of the fishnet unit with respect to that in Ref. 72, the square lattice structure leads to polarization insensitive optical properties at normal incidence. The figure of merit associated with the second-order resonance is noticeably higher than that of the first-order one.70 MP resonance can also be used to combine with gain medium to realize active NIMs at optical frequencies.85–87 A typical fishnet structure was integrated with gain material, as shown in Fig. 3(e), successfully achieving negative refractive index with much improved figure of merit (low loss) in the visible wavelength range between 722 and 738 nm.74 Active medium within the NIM gives rise to an effective gain much higher than its bulk counterpart, which is due to the local-field enhancement inherent in the MP response of NIM.88,89

Except for the applications in NIMs, MP excitation in fishnet structure has also been used for nonlinear metamaterials.90–92 Wang et al. proposed and fabricated a MIM fishnet structure with trilayer Au/MgF2/Au stacks perforated periodically with holes, as demonstrated in Fig. 3(f).75 It supports strong MP resonance due to the excitation of antiparallel currents in the top and bottom metal layers. It was found that the third harmonic radiated from the fishnet structure is a result of the interference of the electric and magnetic dipoles and the electric quadrupole modes.

Except for the above-mentioned fishnet structures mainly used for NIMs, there are many different MIM metamaterials designed with various architectures for other applications.93,94 Early in 2009, a magnetic metamaterial in conjunction with a parallel-metal-plate waveguide [left in Fig. 4(a)] was proposed to realize 2D confinement and guiding of THz waves in the deep subwavelength scale,95 as shown in Fig. 4(a)(top right). A gap plasmon waveguide formed by parallel metal plates can confine light within the deep subwavelength 1D gap.96 This confinement can occur only along the x (electric field) direction due to the fact that it is electric-field-mediated. The beam size of the guided wave is still diffraction limited in the y (magnetic field) direction. On the other hand, in a negative permeability material [bottom right in Fig. 4(a)], the gap can provide the magnetic boundary condition for magnetic-field-mediated confinement, thus supporting MP resonance with the electric field parallel to the material-vacuum interface.97 The MPs excited at each interface are coupled when the gap width w is small enough, which gives rise to symmetric and antisymmetric gap MP modes defined by the Ex symmetry along the y direction. The symmetric mode without cutoff in the zero gap-width limit can be used to confine the light in the y direction, as shown in Fig. 4(b). Moreover, compared with the isotropic effective medium, the magnetic resonator array greatly reduces the angular transmission rate. That is, the scattering loss in the 90° bending geometry, which is caused by defects in the resonator array at the sharp corners of the bend, can be improved by properly arranging the corner resonator array.

FIG. 4.

(a) Schematic diagram of an array of magnetic resonators forming a metamaterial, and (b) its x component of electric field in the y–z plane of waveguides with a 90 bend [(b)-i and (b)-iii] and a splitter [(b)-ii and (b)-iv].95 The claddings in (b)-i and (b)-ii are modeled as an isotropic effective medium, whereas those in (b)-iii and (b)-iv consist of the magnetic resonator array, indicated by the black lines. Reprinted with permission from Ishikawa et al., Phys. Rev. Lett. 102, 043904 (2009). Copyright 2006 American Physical Society. (c) Schematic diagram of the plasmonic sensor with perfect absorbing structure and the incident light polarization configuration, and (d) its calculated antiparallel current distribution when perfect absorbance occurs. (e) Simulated reflectance spectra of the sensor working with water as reference medium.98 Reprinted with permission from Verre et al., Nano Lett. 15, 1952–1958 (2015). Copyright 2015 American Chemical Society. (f) Schematic diagram of the perfect absorber in the x–y plane (i), and the simulated angular dispersions of the absorbance peak with TM (ii) and TE (iii) polarization. (g) Schematic diagram of electrically active switching of magnetic plasmon resonance by way of selective lithium deposition in the lithium metal battery. (h) SEM picture of the original MIM structure (top left), and reflectance spectra before lithium deposition (top right). SEM picture of MIM structure after lithium deposition (bottom left), and reflectance spectra after lithium deposition (bottom right).99 Reprinted with permission from Jin et al., Adv. Mater. 32, 2000058 (2020). Copyright 2020 John Wiley and Sons.

FIG. 4.

(a) Schematic diagram of an array of magnetic resonators forming a metamaterial, and (b) its x component of electric field in the y–z plane of waveguides with a 90 bend [(b)-i and (b)-iii] and a splitter [(b)-ii and (b)-iv].95 The claddings in (b)-i and (b)-ii are modeled as an isotropic effective medium, whereas those in (b)-iii and (b)-iv consist of the magnetic resonator array, indicated by the black lines. Reprinted with permission from Ishikawa et al., Phys. Rev. Lett. 102, 043904 (2009). Copyright 2006 American Physical Society. (c) Schematic diagram of the plasmonic sensor with perfect absorbing structure and the incident light polarization configuration, and (d) its calculated antiparallel current distribution when perfect absorbance occurs. (e) Simulated reflectance spectra of the sensor working with water as reference medium.98 Reprinted with permission from Verre et al., Nano Lett. 15, 1952–1958 (2015). Copyright 2015 American Chemical Society. (f) Schematic diagram of the perfect absorber in the x–y plane (i), and the simulated angular dispersions of the absorbance peak with TM (ii) and TE (iii) polarization. (g) Schematic diagram of electrically active switching of magnetic plasmon resonance by way of selective lithium deposition in the lithium metal battery. (h) SEM picture of the original MIM structure (top left), and reflectance spectra before lithium deposition (top right). SEM picture of MIM structure after lithium deposition (bottom left), and reflectance spectra after lithium deposition (bottom right).99 Reprinted with permission from Jin et al., Adv. Mater. 32, 2000058 (2020). Copyright 2020 John Wiley and Sons.

Close modal

In the application of broadband magnetic metamaterials, plasmonic pentamers constructed from MIM sandwich particles were utilized, successfully realizing a significantly broad magnetic response over an optical range.98 MIM pentamers processes C4 symmetry with four MIM dimers positioned symmetrically around a fifth dimer, as shown in Fig. 4(d). Each vertical MIM dimer behaves as a magnetic meta-atom. It supports an in-phase and an out-of-phase mode. The in-plane magnetic field in the dielectric spacer is produced due to the formation of coil-like currents [Fig. 4(c)]. In the plasmonic Fano pentamer without the atop dielectric caps, destructive interference between a super-radiant mode and a narrow mode produces a dip with an asymmetric Fano-type line shape [Fig. 4(d)]. The combination of the two systems [Figs. 4(c) and 4(d)] forms the MIM oligomer, in which multiple subradiant modes generate a quasi-broadband magnetic response at visible frequencies, as demonstrated in Fig. 4(e).

MP resonances excited by MIM structures can also be used in the application of active plasmonics. Recently, Jin et al. demonstrated the first electrically dynamic control of MP resonance through structure transformation by selective deposition of lithium on a MIM structure.99 The silver-based MIM structure is prepared on the quartz substrate. The double metal strips can be utilized to induce the magnetic dipole moment [top panels in Fig. 4(f)]. When applying a proper electrical deposition current, lithium ions move toward the MIM patterned metasurface through liquid electrolyte [Fig. 4(f)]. The high lithium deposition selectivity of silver can directionally guide the lithium metal depositing vertically along the sidewalls of the MIM.100 This results in structural transformation from MIM to continuous donut geometry. Figure 4(g) gives the SEM images (left) before and after lithium deposition, with the corresponding reflection spectra (right) exhibiting the optical switching from localized magnetic excitations (MP) to delocalized electric excitations (SPP).

Particular interest has also been paid on MP excitations in a special type of MIM structure called nanoparticle-on-mirror (NPoM) system, i.e., a system with nano-sized particles deposited on a metallic film coated with a thin layer of dielectrics.104 The NPoM system exhibits several unique features: (i) simple and practical system with very narrow and tunable gaps which can be controlled within nanometer precision or even subnanometer precision, (ii) easy fabrication and preparation process by chemical synthesis with good extension to large-scale platform, and importantly (iii) excitations of various plasmonic modes and plasmonic-induced magnetic resonance with flexible tunability and strong near-field confinement. All these advantages make the NPoM systems promising candidates for a variety of applications such as optical metamaterials,105 chemical (bio)-sensing,106 and surface-enhanced spectroscopies.103 

A typical MP-assisted metamaterial absorber has been demonstrated based on a NPoM configuration with chemically synthesized silver nanocubes onto a nanoscale-thick polymer spacer layer on a gold film [Fig. 5(a)].101 Each nanocube is the optical analog of a grounded patch antenna (top panel), which supports a series of cavity-like resonances with localized electromagnetic field within the nanoscale gap. For modes with a strongly localized electric field at the patch edges, a magnetic surface current density can be generated in the patch with strong magnetic field confinement, as shown in Fig. 5(b). High efficiency absorption occurs as the total induced effective magnetic current exactly balances the total induced electric current. Such an interferometric effect does not rely on the spatial arrangement of the cubes on the film, thus offering a simple and inexpensive way of obtaining large-area metasurfaces with controlled reflectance.

FIG. 5.

(a) Schematic diagram of the absorber on the basis of NPoM. The enhanced electric field (blue) is equal to effective magnetic currents (red). (b) Reflectance spectra for cubes with specific parameters (top), and the magnetic field excited by the cubes (bottom).101 Reprinted with permission from Moreau et al., Nature 492, 86–89 (2012). Copyright 2012 Springer Nature. (c) Schematic diagram of the plasmonic sensor with perfect absorbing structure and the incident light polarization configuration, and (d) it is calculated antiparallel current distribution when perfect absorbance occurs. (e) Simulated reflectance spectra of the sensor working with water as a reference medium. (f) Schematic diagram of the perfect absorber in the x–y plane (i), and the simulated angular dispersions of the absorbance peak with TM (ii) and TE (iii) polarization.102 Reprinted with permission from Liu et al., Nano Lett. 10, 2342–2348 (2010). Copyright 2010 American Chemical Society. (g) Schematic diagram of gold nanosphere supported by monolayer MBA and Au (top left), schematic diagram of the equivalent LC model of plasmon-induced magnetic dipolar mode (top right), and scattering (black) and absorption (red) spectra for 1 nm and scattering spectra (blue-dashed) for 10 nm dielectric spacer (bottom). (h) Electric field distributions (1-i and 2-i) and magnetic field distributions (1-ii and 2-ii) corresponding to the marked “1” (left panel) and “2” (right panel) in the scattering spectra of (g).103 Reprinted with permission from Chen et al., Nano Lett. 18, 2209–2216 (2018). Copyright 2018 American Chemical Society.

FIG. 5.

(a) Schematic diagram of the absorber on the basis of NPoM. The enhanced electric field (blue) is equal to effective magnetic currents (red). (b) Reflectance spectra for cubes with specific parameters (top), and the magnetic field excited by the cubes (bottom).101 Reprinted with permission from Moreau et al., Nature 492, 86–89 (2012). Copyright 2012 Springer Nature. (c) Schematic diagram of the plasmonic sensor with perfect absorbing structure and the incident light polarization configuration, and (d) it is calculated antiparallel current distribution when perfect absorbance occurs. (e) Simulated reflectance spectra of the sensor working with water as a reference medium. (f) Schematic diagram of the perfect absorber in the x–y plane (i), and the simulated angular dispersions of the absorbance peak with TM (ii) and TE (iii) polarization.102 Reprinted with permission from Liu et al., Nano Lett. 10, 2342–2348 (2010). Copyright 2010 American Chemical Society. (g) Schematic diagram of gold nanosphere supported by monolayer MBA and Au (top left), schematic diagram of the equivalent LC model of plasmon-induced magnetic dipolar mode (top right), and scattering (black) and absorption (red) spectra for 1 nm and scattering spectra (blue-dashed) for 10 nm dielectric spacer (bottom). (h) Electric field distributions (1-i and 2-i) and magnetic field distributions (1-ii and 2-ii) corresponding to the marked “1” (left panel) and “2” (right panel) in the scattering spectra of (g).103 Reprinted with permission from Chen et al., Nano Lett. 18, 2209–2216 (2018). Copyright 2018 American Chemical Society.

Close modal

A infrared perfect plasmonic absorber was also proposed by using a MIM structure consisting of a 2D periodic nanodisk array patterned on a bottom gold mirror separated a thin layer of MgF2 dielectric spacer (several tens of nanometer), as shown in Fig. 5(c).102 Under the illumination of linearly polarized light, antiparallel currents are excited in the gold disk and the bottom gold layer [Fig. 5(d)]. The circulating currents lead to a magnetic moment interacting strongly with the incident light.107,108 This finally gives a distinct MP resonance with nearly perfect absorption due to the strong confinement of electromagnetic energy in the intermediate spacer with no light reflected back. The main feature of this 2D MIM structure [Figs. 5(f)5(i)] lies in the fact that the MP excitation is independent of polarization and angle of incidence, which enables stable and high-efficiency perfect absorption over very large angle for both transverse electric (TE) and transverse magnetic (TM) polarized light [Fig. 5(f)-ii,iii]. The MP resonances are also sensitive to the surrounding dielectric environment. This characteristic was then utilized to realize a plasmonic sensor for refractive index sensing, as shown in Fig. 5(e).

Single particles on a metallic mirror are another type of MIM structure. Delicate control of the particle size and the thickness of the spacer can easily excite multiple resonant electric and magnetic modes with a strong localized electromagnetic field, which is of the potential use for applications in chemical(bio)-sensing and surface-enhanced spectroscopies. A typical NPoM system containing a single nanoparticle was reported in 2018 by Chen et al. as a promising platform for enhanced Raman spectroscopy.103 The proposed system consists of a single gold nanosphere situated on a gold film, spaced by monolayer molecules [Fig. 5(g)]. Under the illumination of polarized light, the single gold nanosphere-film system can induce localized sphere-diameter-dependent electric dipole, quadrupole, and even higher order modes on the sphere surface.109–111 Similar electric modes with antiphase distributions are also excited on the imaging sphere, thus inducing an electric current loop between both particles with the formation of plasmon-induced magnetic resonance (PIMR) under the plasmon-hybridization effect in the sphere-film system. At PIMR resonance [position 1 in the bottom panel of Fig. 5(g)], extreme localization of electric and magnetic fields at the nanogap [left panels in Fig. 5(h)] can be clearly seen. In addition, the current vectors recirculate, indicating an excited magnetic dipolar mode. Two main aspects that contribute to the high capabilities of PIMR on electric and magnetic near-field enhancements: (i) low electromagnetic radiative losses and (ii) efficient coupling between PIMR and both electric- and magnetic-field components of the incident light. Raman spectroscopy has been performed, demonstrating that PIMR significantly enhances the SERS response of probing molecules due to large electric-field enhancements generated by the PIMR.

Metallic arrays of grooves, as one of the most popular nanostructures for SP excitations, have been widely used for both fundamental research such as optical nonlinearity as well as light–matter interactions, and a variety of practical applications for the design of plasmonic and nanophotonic devices such as plasmonic waveguides, biosensors optical absorbers and modulators.24,112–119

Among the groove structures, a special type of structure called “deep grooves,” i.e., grooves with large depths and high aspect ratio, have drawn much interest due to their capability to supporting MP resonances under certain conditions of illumination.121–127 MP resonances in deep groove structures can strongly confine both electric and magnetic fields in the nano-sized groove gaps, thus offering ideal platforms for energy harvesting and sensing applications.

A typical deep groove array, as demonstrated Fig. 6(a) (top), was theoretically proposed to generate MP resonance to enhance the absorption in graphene in near-infrared region.120 It is well known that Graphene, as a two-dimensional monatomic layer of carbon material, exhibits unique electric and optical properties.128 Due to its high carrier mobility of graphene at room temperature, graphene has been demonstrated as a good candidate for ultrafast optoelectronic devices such as transistors129,130 and photodetectors.131,132 However, graphene is essentially transparent in the visible and near-infrared with an absorptivity of 2.3%. This greatly hinders its application in photon detection,133 optical antennas134 and solar cells.135 To improve absorption in graphene, a 1D sliver grating was proposed by Zhao et al. to combine with a monolayer of graphene. With TM waves at normal incidence, the deep grooves (200 nm) with small trench openings (30 nm) can support firs-order (MP1 with wavelength λ=1.49μm) and second-order (MP2 with wavelength λ=541 nm) MP resonances [middle panel in Fig. 6(a)]. The MP resonances originate from a circulating current loop [bottom panel in Fig. 6(a)] induced by the magnetic field of the incident wave in the deep grooves. A strong magnetic field is confined in the trenches for the structure with and without graphene. Importantly, the electric field near the opening of the trench is enhanced to about 20 times greater than that of the incidence in the case of graphene-covered grating, which enables the graphene to absorb most of the incident power. As a result, in the presence of graphene monolayer, the overall absorptance increases from 0.21 to 0.81 at MP1 and from 0.62 to 0.99 at MP2, without changing the MP resonance frequencies [middle panel in Fig. 6(a)].

FIG. 6.

(a) Schematic diagram of the deep Ag grating covered with graphene (top), and absorptance spectra of the grating with (blue solid line) and without (red dashed line) graphene (middle), and electromagnetic field for grating without (bottom left) and with (bottom right) graphene at the resonance.120 Reprinted with permission from Zhao et al., Appl. Phys. Lett. 105, 031905 (2014). Copyright 2014 AIP Publishing LLC. (b) Schematic diagram and (c) SEM picture of the graphene supported by the gold groove array (d) Experimental (solid lines) and simulated (dashed lines) reflectance spectra of the groove with (red lines) and without (blue lines) graphene. (e) The simulation of the distribution of the amplitude of the magnetic field inside the groove without (left) and with (right) graphene.121 Reprinted with permission from Wang et al., Phys. Rev. B 99, 235407 (2019). Copyright 2019 American Physical Society. (f) Schematic diagram (i), top view SEM picture (ii), and side view SEM picture of gold groove array. (iv) Electric field at MP resonance. (g) Measured reflectance spectra of the sensor with different refractive indices.122 Reprinted with permission from Li et al., Opt. Express 26, 34122–34130 (2018). Copyright 2018 The Optical Society.

FIG. 6.

(a) Schematic diagram of the deep Ag grating covered with graphene (top), and absorptance spectra of the grating with (blue solid line) and without (red dashed line) graphene (middle), and electromagnetic field for grating without (bottom left) and with (bottom right) graphene at the resonance.120 Reprinted with permission from Zhao et al., Appl. Phys. Lett. 105, 031905 (2014). Copyright 2014 AIP Publishing LLC. (b) Schematic diagram and (c) SEM picture of the graphene supported by the gold groove array (d) Experimental (solid lines) and simulated (dashed lines) reflectance spectra of the groove with (red lines) and without (blue lines) graphene. (e) The simulation of the distribution of the amplitude of the magnetic field inside the groove without (left) and with (right) graphene.121 Reprinted with permission from Wang et al., Phys. Rev. B 99, 235407 (2019). Copyright 2019 American Physical Society. (f) Schematic diagram (i), top view SEM picture (ii), and side view SEM picture of gold groove array. (iv) Electric field at MP resonance. (g) Measured reflectance spectra of the sensor with different refractive indices.122 Reprinted with permission from Li et al., Opt. Express 26, 34122–34130 (2018). Copyright 2018 The Optical Society.

Close modal

This theoretical idea was verified by a detailed experiment in 2019.121 Wang et al. fabricated a simple 1D gold groove array with a monolayer of graphene deposited on top [Figs. 6(b) and 6(c)]. The strong interaction of graphene with the MP resonance was intensively studied using linear static spectroscopy. Reflectivity measurement and numerical simulation clearly demonstrated that MP resonance was excited around 1150 nm, with a strong magnetic field localized inside the groove [Fig. 6(e)]. At MP resonance, the graphene has an enhanced absorption by one order of magnitude with respect to the 2.3% light absorption for the pristine graphene monolayer, as shown in Fig. 6(d). Interestingly, the strong graphene-MP interaction also suppresses the Ohmic loss in the metallic groove array. In contrast to other reported graphene-plasmonic systems with considerable resonance shift, limited metal-graphene charge transfer at the directly contacting interface leads to a reduced Fermi level of graphene without introducing an additional shift of magnetic resonances. The observed considerable MP resonance shift essentially results from the small variation of the dielectric environment inside the grooves.

The strongly localized electric and magnetic field in the deep grooves at MP resonance also provides a promising way of realizing sensing devices. Zhu et al. proposed a similar 1D deep groove array for refractive index sensing.136 A sensitivity up to 1200 nm/RIU was achieved with a figure of merit of 15, which benefited from the MP resonances that are extremely sensitive to the surrounding media. The proposed structure was later experimentally fabricated, as shown in Fig. 6(f)i–iii. The sensing device consists of a periodic nanogroove array featured by deep grooves with a high aspect ratio. Numerical simulation confirms the excitation of MP resonance, at which a circulating current is formed around the groove with strong localization of the magnetic field inside the groove [Fig. 6(f)-iv]. Utilizing MP resonances that are extremely sensitive to the surrounding dielectric medium [Fig. 6(g)], a refractive index sensitivity up to 1300 nm/RIU was achieved in the near infrared region. Importantly, benefiting from angle-independent MP resonances with strong confinement of the magnetic field inside the deep grooves and strong electric field localization at the groove openings, it exhibits wide-angle sensing capability with an excellent linear dependence on the change of the refractive index.

Plasmonic nanoclusters, i.e., nanoparticles placed in close proximity with different configurations, are another type of architecture that hold great promise for artificial magnetic phenomena.19,20,137–141 Similar to the above-mentioned MIM and SRR structures which resemble LC elements, plasmonic nanoclusters are analogous to molecules formed by the arrangement of atoms interconnected by chemical bonds. These assembly of nanoparticles are usually called metamolecules. When brought into close proximity, plasmons induced by the individual nanoparticles start to couple with each other, leading to collective hybridized modes and even higher order electric and MP modes, similar to molecular orbitals.142–145 This analogy is, therefore, extremely convenient in describing the inter-particle coupling. Just as the properties of a molecule are defined by the relative position of atoms, the optical property of plasmonic nanoclusters are determined by the relative position of each object.146,147Figure 7(a) shows a series of typical nanoclusters with self-assembled nanoshells as the basic building blocks.146 The nanoshell contains a silica core and gold shell, which is coated further with a polymer to generate nanogaps between nanoparticles and control the magnitude of interparticle electromagnetic coupling. For individual nanoshells and coupled dimers, electric resonances dominate under the illumination of a vertically polarized light [left panel in Fig. 7(a)]. The resonance shifts to red with decreasing gap separation because of enhanced capacitive coupling between the nanoshells. While for the trimer consisting of three nanoshells of equilateral spacing, magnetic dipole resonances are clearly visible in the scattering spectrum, which can be described as a closed loop of nanoinductors and nanocapacitors [middle panel in Fig. 7(a)]. Evidence for a magnetic dipole mode comes from spectra taken with the cross-polarizer in the light path, demonstrating a clear narrow magnetic dipole peak around 1400 nm [Fig. 7(b)]. For heptamers, symmetric clusters composed of seven equivalent elements, complex plasmon mode interactions result in Fano-like spectral line shape [right panel in Fig. 7(a)]. In the measured scattering spectra given in Fig. 7(c), there exists a bright mode (1160 nm) resulting from charge oscillations in each nanoshell oriented in the same direction due to constructive interference of the radiating fields. While the charge oscillations oriented in different directions give rise to the other dark mode at 1490 nm due to the destructive interference of the radiating fields.

FIG. 7.

(a) Different resonances excited by clusters consists of a different number of nanoshells. (b) Experimental s-polarized scattering spectra with insertion of the cross-polarizer for an individual trimer shown in the inset. (c) TEM picture and spectra of a heptamer at three different incident electric field orientation angles.146 Reprinted with permission from Capasso et al., Science 328, 1135–1138 (2010). Copyright 2010 The American Association for the Advancement of Science. (d) Schematic diagram of AFM nanomanipulation and optical scattering set-up with the asymmetric nanoring. (e) AFM pictures of the nanorings. (f) Measured scattering spectra under s-polarized incident light when the four nanoparticles are far apart (left) and when they are arranged with small gaps (right).148 Reprinted with permission from Shafiei et al., Nat. Nanotechnol. 8, 95–99 (2013). Copyright 2013 Springer Nature.

FIG. 7.

(a) Different resonances excited by clusters consists of a different number of nanoshells. (b) Experimental s-polarized scattering spectra with insertion of the cross-polarizer for an individual trimer shown in the inset. (c) TEM picture and spectra of a heptamer at three different incident electric field orientation angles.146 Reprinted with permission from Capasso et al., Science 328, 1135–1138 (2010). Copyright 2010 The American Association for the Advancement of Science. (d) Schematic diagram of AFM nanomanipulation and optical scattering set-up with the asymmetric nanoring. (e) AFM pictures of the nanorings. (f) Measured scattering spectra under s-polarized incident light when the four nanoparticles are far apart (left) and when they are arranged with small gaps (right).148 Reprinted with permission from Shafiei et al., Nat. Nanotechnol. 8, 95–99 (2013). Copyright 2013 Springer Nature.

Close modal

Similar nanocluster architecture, with four closely packed nanoparticles arranged in a ring configuration [Figs. 7(d) and 7(e)], was later on reported by Shafiei et al., demonstrating the manipulation of the electric and magnetic response by controlling the inter-particle separations.148 When placed nanorings far apart [left panel in Fig. 7(e)], a single, broad resonance arising from the dipolar contribution of the individual nanorings is visible in the scattering spectrum [left panel in Fig. 7(f)]. As the inter-particle gaps decreases down to sub-10 nm [right panel in Fig. 7(e)], the scattering spectrum experiences a red-shift of the electric dipole resonance due to an increase of the capacitance induced by the mutual inter-particle coupling. The reduction of the interparticle gap allows the efficient formation of LC circuits: an effective hybridization between the radiative (wide) electric dipole and the subradiative (narrow) magnetic modes, ultimately resulting in a pronounced Fano skew in the scattered spectrum [right panel in Fig. 7(f)].

Owning to the capability of manipulating mutual particle couplings for the generation of tunable optical property associated with MP resonances, metamolecules have found their applications in many fields such as plasmonic waveguides,22,149 antenna,151 energy harvesting,21 biodetection,23 and non-linear light generation.140,152

Using a heptamer structure with planar arrangements of fused ring patterns, Liu et al. have taken advantage of antiphase MPs excited in adjacent fused heptamer units to generate plasmonic “antiferromagnetic” behavior in multiple repeated heptamer structures, successfully realizing a low-loss MP waveguide.149 The plasmonic heptamer serves as a benzene-analog building block for the construction of cyclic aromatic structures, as shown in Fig. 8(a). In each heptamer, ring currents are excited circulating around the six nanoparticles. Two antiparallel magnetic dipole moments are, therefore, excited by the ring current flowing through the shared junction. For excitation of the structure with linearly polarized light parallel to the direction of the two shared gold nanoparticles at normal incidence, the measured extinction spectrum exhibits two distinct subradiant resonances, as shown in Fig. 8(b)(left). Numerical simulation [right panel in Fig. 8(b)] very nicely reproduced the experimental results. Resonance I around 1100 nm is a double Fano resonance mode with antiphase charge oscillation in the two center particles relative to those in the surrounding particles (bottom inset). While resonance II at 1550 nm is featured by ring currents circulating in opposite directions around the two fused rings (middle inset). In this case, the two shared gold nanoparticles functioning as a mutual current link (top inset), thus forming the excitation of two antiparallel magnetic dipole moments in the plasmonic naphthalene structure.

FIG. 8.

(a) SEM picture of a plasmonic naphthalene sample. The inset shows an enlarged view of the sample. (b) Experimental and simulated extinction spectra of the sample. Insets show charge density and magnetic field at resonances I and II. (c) Charge density picture of a plasmonic aromatic chain with twelve heptamer units excited by a dipole excitation source.149 Reprinted with permission from Liu et al., Nano Lett. 12, 364–369 (2012). Copyright 2012 American Chemical Society. (d) Simulated extinction efficiency spectra of the disk trimer (red curve) and of the moon-integrated trimer (black curve), respectively. For clarity reasons, the red curve is shifted with respect to the black one. The insets show the current density distribution around the m dip. (e) 2D pictures of the magnetic field enhancement distribution at Fano resonance for the disk trimer and the moon trimer resonator (MTR). (f) SEM pictures (left column), and 2D magnetic field enhancement distributions at coil-type Fano resonance (right column) of MTR with different moon apical gaps. (g) The spectral evolution of the squeezing factor for nanodisk trimers (red dots) and MTRs (black dots), respectively (empty dots: simulation, full dots: experiment). The inset shows the relation between the apical gap G and the resonance wavelength of MTRs.21 Reprinted with permission from Panaro et al., Nano Lett. 15, 6128–6134 (2015). Copyright 2015 American Chemical Society. (h) Schematic picture and parameters of the disk trimer supporting the magnetic Fano resonance (top), and scattering spectra of magnetic Fano resonance in the trimer (bottom). The inset shows the magnetic dipole in the trimer structure. (i) The SHG in-plane electric near field (left) and far field (right) in the central plane of the trimer at the wavelength of 450 nm. E0 is the electric field of the incident light.150 Reprinted with permission from Yang et al., Nano Lett. 9, 6068–6075 (2017). Copyright 2017 The Royal Society of Chemistry.

FIG. 8.

(a) SEM picture of a plasmonic naphthalene sample. The inset shows an enlarged view of the sample. (b) Experimental and simulated extinction spectra of the sample. Insets show charge density and magnetic field at resonances I and II. (c) Charge density picture of a plasmonic aromatic chain with twelve heptamer units excited by a dipole excitation source.149 Reprinted with permission from Liu et al., Nano Lett. 12, 364–369 (2012). Copyright 2012 American Chemical Society. (d) Simulated extinction efficiency spectra of the disk trimer (red curve) and of the moon-integrated trimer (black curve), respectively. For clarity reasons, the red curve is shifted with respect to the black one. The insets show the current density distribution around the m dip. (e) 2D pictures of the magnetic field enhancement distribution at Fano resonance for the disk trimer and the moon trimer resonator (MTR). (f) SEM pictures (left column), and 2D magnetic field enhancement distributions at coil-type Fano resonance (right column) of MTR with different moon apical gaps. (g) The spectral evolution of the squeezing factor for nanodisk trimers (red dots) and MTRs (black dots), respectively (empty dots: simulation, full dots: experiment). The inset shows the relation between the apical gap G and the resonance wavelength of MTRs.21 Reprinted with permission from Panaro et al., Nano Lett. 15, 6128–6134 (2015). Copyright 2015 American Chemical Society. (h) Schematic picture and parameters of the disk trimer supporting the magnetic Fano resonance (top), and scattering spectra of magnetic Fano resonance in the trimer (bottom). The inset shows the magnetic dipole in the trimer structure. (i) The SHG in-plane electric near field (left) and far field (right) in the central plane of the trimer at the wavelength of 450 nm. E0 is the electric field of the incident light.150 Reprinted with permission from Yang et al., Nano Lett. 9, 6068–6075 (2017). Copyright 2017 The Royal Society of Chemistry.

Close modal

Based on this principle, efficient MP propagation along a chain of twelve heptamer units was demonstrated using a dipole excitation source, as shown in Fig. 8(c). With the help of strong near-field coupling, MPs can propagate along the chain via the excitation of alternating ring currents in adjacent heptamers, leading to “antiferromagnetic” MP behavior with the induced magnetic moments aligned in an antiparallel manner with respect to each other along the entire multiple heptamer chain.

With the proper arrangement of metallic nanoparticles, MP resonances can also be utilized to squeeze light for nanofocusing applications. Panaro et al. proposed and fabricated two special types of plasmonic nanoclusters, demonstrating a strong magnetic field nanofocusing in the infrared.21 The plasmonic nanoclusters, as shown in the insets of Fig. 8(d), consist of disk and moon trimer resonator (MTR). When illuminated by a linearly polarized light at normal incidence, the simulated extinction efficiency spectra of both trimers show clear dips (denoted as m), but the moon trimer exhibits a narrower spectral shape at m point with a intensified magnetic hot-spot due to the excitation of quardupole-like MP resonance with respect to the dipolar peak of the large disk trimer. These dips in both cases correspond to a coil-type current distribution in the gap region of the clusters with a magnetic hot-spot in the center of the gap [Fig. 8(e)]. The advantage of the moon trimer lies in the following two aspects: (i) the MTR provides a much higher squeezing factor compared to the disk counterpart [black line in Fig. 8(g)], and importantly (ii) one can readily tune the apical gaps of the moon to control the position of the MP resonances [inset of Fig. 8(f)] while keeping a high-level of magnetic nanofocusing capability, with no need to change the size of the coil high-magnetic-field region [right in Fig. 8(f)]. These features enable a straightforward boost of the enhancement of the infrared magnetic field with efficient squeezing in localized nanovolumes.

In addition to the applications in the linear spectral regime, MP excitations in plasmonic nanoclusters can also be used to generate the magnetic Fano resonance-induced nonlinearity enhancement for nonlinear light generation.150,152 Yang et al. proposed plasmonic metamolecule rings to enhance second-harmonic generation (SHG).150 As shown in Fig. 8(h), a trimer structure contains one bigger nanodisk accompanied by two smaller ones. The inter-particle gaps were rationally determined for obtaining the strongest Fano resonance while considering the practical fabrication possibility.153 Simulated scattering cross section exhibits a very sharp Fano dip around 900 nm, which is not observed in the absorption cross section [bottom panel in Fig. 8(h)]. This dip results from the interference of the superradiant electric mode with the subradiant magnetic mode. When the electric dipole moments of each nanodisk are in the same direction, the superradiant electric mode is formed with a broad peak appearing in the optical spectra. Ring displacement currents produce the subradiant magnetic mode when the electric dipole moments are arranged end-to-end. In this case, the electric displacement field forms a complete circle with a strong magnetic field perpendicular to the plane in the gap region [the inset of Fig. 8(h)]. A type of Fano resonance is generated due to the subradiant magnetic mode interferes with the broad superradiant mode. This Fano resonance is relevent to the magnetic field, therefore, it is also called magnetic Fano resonance. The key feature of magnetic Fano resonance is the minimum energy radiated from the nano oscillators, which enables more energy localized at the surface of the nanostructure with a strong fundamental near field. Since the trimer suffers lower radiative losses and produces a stronger near field at the Fano dip, it facilitates the generation of a stronger SHG signal at the Fano dip. Figure 8(i) shows the SHG near field (left) and far field (right) at 450 nm. Consistent with the near field at the fundamental frequency. The SHG near field for the excitation at the Fano dip shows stronger intensity than that at the scattering peaks. The capability of such MP-based nanoclusters holds great promise for the manipulation of nonlinear optical processes with high efficiency.

As reviewed above, magnetic nanostructures such as deep groove arrays and NPoM systems can strongly confined the incident light to form intense magnetic hotspots at MP resonances, which are of great help in dramatically boosting light–matter interactions in nanoscale volumes, with remarkable enhancement of absorption in graphene,120,121 boosted SHG,36 as well as enhanced Raman signals.103 Magnetic nanostructures can also be designed to realize even stronger light–matter interactions by bringing MP resonances to couple with other optical excitations, such as photonic resonances,154 optical cavity modes,155 delocalized plasmon modes156 and excitonic transitions in 2D materials.157–159 Particular interest has been drawn in MP-involved interactions with excitonic transitions in a strong coupling regime, in which magnetic polaritons containing the characteristics of both MP and excitons offer ideal platforms for the study of light–matter interactions, as well as for the potential applications in realizing novel quantum and nanophotonic optical devices.157,159

Early in 2006, Linden et al. have studied the strong coupling between MP resonance and a Bragg resonance in 1D magnetic photonic crystals [top right panel in Fig. 9(a)], which consists of a period array of magnetic atoms on top of a dielectric slab waveguide.154 The magnetic atoms [left panel in Fig. 9(a)] are pairs of long gold wires separated by a thin layer of magnesium fluoride (MgF2). For TM-polarized incident light perpendicular to the wire pairs, the magnetic atoms exhibit two pronounced resonances in the spectral region of interest. The long-wavelength resonance corresponds to MP resonance, i.e., a strong magnetic-dipole moment with the magnetic field is concentrated in the MgF2 spacer oriented along the wires. The MP resonance is related to the antisymmetric eigenmode with the electric fields pointing in opposite directions in the two wires. A quasiguided TM-waveguide mode is excited for normal incidence due to the periodic arrangement of wire pairs. By increasing the lattice constant, the Bragg mode shifts to a longer wavelength. As the two modes are brought into resonance, strong coupling occurs with a clear anti-crossing behavior in both the measured [Fig. 9(b)] and simulated transmission spectra [Fig. 9(c)]. The Bragg and the magnetic resonance are detuned with increasing lattice constants, which basically yields an uncoupled system again.

FIG. 9.

(a) Schematic diagram (left) and top view electron micrograph (bottom right) of gold-wire pairs supported by a dielectric waveguide, and schematic diagram of a one-dimensional magnetic photonic crystal. (b) Measured and (c) simulated transmittance for (a) on an gray scale vs lattice constant and wavelength.154 Reprinted with permission from Linden et al., Phys. Rev. Lett. 97, 083902 (2006). Copyright 2006 American Physical Society. (d) Schematic diagram of a NPoM-TMD strong coupling system (left), and the magnetic field distribution of the pure NPoM system (right). (e) Simulated normalized scattering spectra of the hybrid system as a function of rod length. (f) The time-domain evolution of the exciton (blue) and plasmon-induced magnetic resonance (red) populations.157 Reprinted with permission from Xie et al., Phys. Rev. B 101, 045403 (2020). Copyright 2020 American Physical Society. (g) Schematic diagram of a Sb2Te3 topological insulators nanogroove (left), magnetic field distribution of the nanogroove unit at resonance wavelength (right). (h) Photoluminescence (PL) emission spectra from the pure nanogrooves and the film/nanogrooves covered by monolayer MoS2 (left), enhancement factors of the electric field in nanogrooves as a function of polarization angle (right). The inset shows the PL intensity spectra when the polarization angle is zero.158 Reprinted with permission from Lu et al., Light Sci. Appl. 9, 191 (2020). Copyright 2020 Springer Nature.

FIG. 9.

(a) Schematic diagram (left) and top view electron micrograph (bottom right) of gold-wire pairs supported by a dielectric waveguide, and schematic diagram of a one-dimensional magnetic photonic crystal. (b) Measured and (c) simulated transmittance for (a) on an gray scale vs lattice constant and wavelength.154 Reprinted with permission from Linden et al., Phys. Rev. Lett. 97, 083902 (2006). Copyright 2006 American Physical Society. (d) Schematic diagram of a NPoM-TMD strong coupling system (left), and the magnetic field distribution of the pure NPoM system (right). (e) Simulated normalized scattering spectra of the hybrid system as a function of rod length. (f) The time-domain evolution of the exciton (blue) and plasmon-induced magnetic resonance (red) populations.157 Reprinted with permission from Xie et al., Phys. Rev. B 101, 045403 (2020). Copyright 2020 American Physical Society. (g) Schematic diagram of a Sb2Te3 topological insulators nanogroove (left), magnetic field distribution of the nanogroove unit at resonance wavelength (right). (h) Photoluminescence (PL) emission spectra from the pure nanogrooves and the film/nanogrooves covered by monolayer MoS2 (left), enhancement factors of the electric field in nanogrooves as a function of polarization angle (right). The inset shows the PL intensity spectra when the polarization angle is zero.158 Reprinted with permission from Lu et al., Light Sci. Appl. 9, 191 (2020). Copyright 2020 Springer Nature.

Close modal

In contrast to the 1D multi-stacked magnetic photonic crystals, the NPoM system supporting MPs has recently drawn much attention for the study of light–matter interactions in strong coupling regime due to its compact structural configuration providing ultrasmall mode volumes.157,159 Xie et al. proposed a hybrid NPoM system consisting of a single silver nanorod situated on a gold film separated by a thin Al2O3 spacer with an atomic-thick tungsten disulfide (WS2) monolayer embedded between the rod and the dielectric layer, as depicted in Fig. 9(d). Under the illumination of the incident light with the magnetic field perpendicular to the rod, the bare NPoM system with nanorod (without the WS2 monolayer) can support a strong MP mode (namely, PIMR defined in Sec. C). The underlying mechanism of PIMR excitation is analogous to that of a NPoM system with a nanosphere:103 one pair of antiparallel currents in the top nanorod and its mirror image in the silver film is induced by the oscillating magnetic field. These two currents cause a diamagnetic response,122,125,148 which produces an electric current loop between real and imaging rods with the excitation of MP modes under the plasmon hybridization effect.101,103 The strong MP resonance can be confirmed by the simulated local magnetic field distributions at the peak wavelength, as shown in the right panel of Fig. 9(d). A circulating current loop (red arrows) is produced around the nanorod and its mirror image with a strong localized magnetic field inside the spacer. Due to the strongly localized magnetic field, the corresponding electric field is distributed mainly inside the dielectric nanocavity and at the bottom edges of the hemispheres, which plays an important role in achieving strong MP-exciton coupling with large Rabi splitting. In the hybrid NPoM system in presence of WS2 monolayer, the MP resonance can be tuned across the exciton transition energy of the WS2 monolayer by simply increasing the rod length. In the simulated scattering spectra [Fig. 9(e)], a typical anticrossing behavior can be apparently seen in the 2D color map with two clear split branches (dashed black), indicating the strong coherent MP-exciton interaction and the formation of a hybrid (energetically) upper polariton (UP) and lower polariton (LP). A large Rabi splitting over 220 meV was obtained of excitons involved in the coupling. Both the large splitting and the small mode volume led to only a few number of excitons participating in the coupling. A full quantum model is established to quantitatively describe the coherent and incoherent coupling dynamics of the hybrid modes. An ultrafast energy exchange with a period of 19 fs can be clearly visualized in the population dynamics of the LP branch [Fig. 9(f)].

MP can also be used to couple with 2D materials for the enhancement of spontaneous emission, in this case, the coupling strength between the two subsystems is not large enough to reach a strong coupling regime, instead, they are in the weak coupling regime. Recently, strongly enhanced light-MoS2 interaction has been demonstrated by using MP resonances excited in nanostructured topological insulators (TIs).158 As novel states of matter, TI presents topologically protected conducting surfaces and insulating bulks in a broad optical range, providing new building blocks for plasmonics. MP resonances were demonstrated for the first time by Lu et al. in TI structures composed of nanofabricated Sb2Te3 nanogrooves, as shown in Fig. 9(g). The Sb2Te3 material possesses distinctly discrepant surface and bulk states, exhibiting excellent optical characteristics.160,161 The metal-like property of the surface state provides the possibility of generating the plasmonic response at high frequencies.142,162,163

When illuminated by the incident light with polarization perpendicular to the nanogrooves, a distinct reflection dip occurs at an optical frequency with wavelengths around 736 nm. Simulated near field distribution reveals the formation of MP resonance with a strong magnetic field concentrated inside the groove [left panel in Fig. 9(g)], exhibiting a strong diamagnetic effect. Such MP resonance in the TI nanostructure with strong field enhancement is expected to offer a promising route for boosting light–matter interactions. By integrating a MoS2 atomic monolayer with the TI grooves, the photoluminescence (PL) emission of monolayer MoS2 based on MPR is greatly enlarged with respect to the case for bare TI films [left panel in Fig. 9(h)]. Importantly, the PL enhancement factor can be readily tuned by changing the polarization angle of the incident light [right panel in Fig. 9(h)]. The enhancement factor of the electric field intensity is close to the reinforced strength of the MoS2 PL excitonic peak. Therefore, electric field enhancement plays a critical role in PL enhancement.

In this Perspective, we comprehensively review the current studies on MPs excited in plasmonic nanostructures and their applications in many interdisciplinary areas of research. MP resonances, initially excited in SRR-based metamaterials for achieving artificial magnetism, have attracted much attention over the last two decades. With the development of nanofabrication techniques, MP resonances have been successfully pushed from microwave to visible range by employing nano-sized complex architectures such as plasmonic oligomers. This offers powerful platforms for the manipulation of magnetism in the desired wavelength range, thus offering promising approaches to a variety of applications such as light-harvesting, emission control, biodetection, and non-linear light generation.

Although many advances and achievements have been obtained in the aforementioned aspects, there are still new possible directions in the future. Coupling of MP resonances to other optical excitations including plasmonic cavity modes, surface lattice resonances, or excitonic emitters in novel 2D materials may give rise to polaritonic states with fascinating physical properties. Particularly, strong coupling between MP and excitons in 2D materials has attracted much attention from researchers. Ultrafast dynamics in such strongly coupled systems is becoming an important research topic. Comprehensive studies on the microscopic MP-exciton coupling dynamic may be of great help for a deep understanding of the strong light–matter interaction, as well as for the potential applications in realizing novel quantum and nanophotonic optical devices. Challenges still exist for the studies on MPs supported in plasmonic nanostructures. Particular attention has been paid to finding ways for both flexible manipulation at the nanoscale and the suppression of the inherent losses suffered by circulating currents in plasmonic materials. In this sense, all-dielectric nanoparticles made of high-index semiconducting materials have been proposed due to their capability of supporting both magnetic and electric Mie resonances with extremely low loss. Compared with their conventional metallic counterparts, the dielectric structures are endowed with weaker near field confinement. This disadvantage may hinder their applications in the fields for which strong light–matter interactions are required. We expect that integrating plasmonic materials with all-dielectric structures may combine the ability of plasmonic materials in strong field confinement and flexible tunability of MP excitations in dielectric structures. This is expected to pave the way toward the realization of new platforms for the study of artificial magnetic materials.

This work was supported by the National Natural Science Foundation of China (Grant Nos'. 11974254 and 11974253), Science Speciality Program of Sichuan University (Grant No. 2020SCUNL210), and the Innovation Program of Sichuan University (Grant No. 2018SCUH0074).

The authors have no conflicts to disclose.

Yuyang Wu: Writing – original draft (equal); Writing – review & editing (equal). Peng Xie: Writing – original draft (equal); Writing – review & editing (equal). Qi Ding: Writing – original draft (equal); Writing – review & editing (equal). Yuhang Li: Writing – original draft (supporting); Writing – review & editing (equal). Ling Yue: Writing – original draft (supporting); Writing – review & editing (equal). Hong Zhang: Supervision (equal). Wei Wang: Project administration (equal); Supervision (equal).

Data sharing is not applicable to this article as no new data were created or analyzed in this study.

1.
L. D.
Landau
and
E. M.
Lifshitz
,
Electrodynamics of Continuous Media
(
Wiley
,
New York
,
1984
).
2.
L. D.
Landau
and
E. M.
Lifshitz
,
Electrodynamics of Continuous Media Section 60
(
Pergamon
,
Oxford
,
1960
).
3.
Z.
Han
and
S. I.
Bozhevolnyi
,
Rep. Prog. Phys.
76
,
016402
(
2013
).
4.
J.
Tong
,
F.
Suo
,
J.
Ma
,
L. Y. M.
Tobing
,
L.
Qian
, and
D. H.
Zhang
,
Opto-Electron. Adv.
2
,
180026
(
2019
).
5.
J.
Butet
,
P. F.
Brevet
, and
O. J. F.
Martin
,
ACS Nano
9
,
10545
10562
(
2015
).
6.
E.
You
,
Y.
Chen
,
J.
Yi
,
Z.
Meng
,
Q.
Chen
,
S.
Ding
,
H.
Duan
,
M.
Moskovits
, and
Z.
Tian
,
Opto-Electron. Adv.
4
,
210076
(
2021
).
7.
K.
Bae
,
G.
Kang
,
S. K.
Cho
,
W.
Park
,
K.
Kim
, and
W. J.
Padilla
,
Nat. Commun.
6
,
10103
(
2015
).
8.
X.
Wen
,
Y.
Bi
,
F.
Yi
,
X.
Zhang
,
Y.
Liu
,
W.
Wang
, and
J.
Feng
,
Opto-Electron. Adv.
4
,
200024
(
2021
).
9.
10.
A.
Gonzalez-Tudela
,
D.
Martin-Cano
,
E.
Moreno
,
L.
Martin-Moreno
,
C.
Tejedor
, and
F. J.
Garcia-Vidal
,
Phys. Rev. Lett.
106
,
020501
(
2011
).
11.
S.
Jiao
,
J.
Liu
,
L.
Zhang
,
F.
Yu
,
G.
Zuo
,
J.
Zhang
,
F.
Zhao
,
W.
Lin
, and
L.
Shao
,
Opto-Electron. Sci.
1
,
220010
(
2022
).
12.
F.
Monticone
and
A.
Alu
,
J. Mater. Chem. C
2
,
9059
9072
(
2014
).
13.
G. T.
Papadakis
,
D.
Fleischman
,
A.
Davoyan
,
P.
Yeh
, and
H. A.
Atwater
,
Nat. Commun.
9
,
296
(
2018
).
14.
J. B.
Pendry
,
A. J.
Holden
,
D. J.
Robbins
, and
W. J.
Stewart
,
IEEE T. Microw. Theory
47
,
2075
2084
(
1999
).
15.
R. A.
Shelby
,
D. R.
Smith
, and
S.
Schultz
,
Science
292
,
77
79
(
2001
).
16.
G.
Dolling
,
C.
Enkrich
,
M.
Wegener
,
C. M.
Soukoulis
, and
S.
Linden
,
Science
312
,
892
894
(
2006
).
17.
J.
Zhou
,
T.
Koschny
,
M.
Kafesaki
,
E. N.
Economou
,
J. B.
Pendry
, and
C. M.
Soukoulis
,
Phys. Rev. Lett.
95
,
223902
(
2005
).
18.
Y.
Ren
,
Y.
Lu
,
T.
Zang
,
Y.
Wang
,
Y.
Dai
, and
P.
Wang
,
Opt. Express
25
,
28417
28426
(
2017
).
19.
P.
Wang
,
J. H.
Huh
,
J.
Lee
,
K.
Kim
,
K. J.
Park
,
S.
Lee
, and
Y.
Ke
,
Adv. Mater.
31
,
1901364
(
2019
).
20.
F.
Wen
,
J.
Ye
,
N.
Liu
,
D. P. V.
P. Nordlander
, and
N. J.
Halas
,
Nano Lett.
12
,
5020
5026
(
2012
).
21.
S.
Panaro
,
A.
Nazir
,
R. P.
Zaccaria
,
L.
Razzari
,
C.
Liberale
,
F. D.
Angelis
, and
A.
Toma
,
Nano Lett.
15
,
6128
6134
(
2015
).
22.
N.
Liu
,
S.
Mukherjee
,
K.
Bao
,
Y.
Li
,
L. V.
Brown
,
P.
Nordlander
, and
N. J.
Halas
,
ACS Nano
6
,
5482
5488
(
2012
).
23.
I.
Jung
,
S.
Ih
,
H.
Yoo
,
S.
Hong
, and
S.
Park
,
Nano Lett.
18
,
1984
1992
(
2018
).
24.
S.
Zhang
,
H.
Zhang
,
T.
Xu
,
W.
Wang
,
Y.
Zhu
,
D.
Li
,
Z.
Zhang
,
J.
Yi
, and
W.
Wang
,
Phys. Rev. B
97
,
235401
(
2018
).
25.
W.
Du
,
Y.
Zhu
,
Z.
Yan
,
X.
Xu
,
X.
Xu
,
J.
Hu
,
P.
Cai
, and
C.
Tang
,
Nanomaterials
11
,
2194
(
2021
).
26.
R. M.
Bakker
,
D.
Permyakov
,
Y. F.
Yu
,
D.
Markovich
,
R.
Paniagua-Dominguez
,
L.
Gonzaga
,
A.
Samusev
,
Y.
Kivshar
,
B.
Luk’yanchuk
, and
A. I.
Kuznetsov
,
Nano Lett.
15
,
2137
2142
(
2015
).
27.
A. E.
Miroshnichenko
and
Y. S.
Kivshar
, “
Fano resonances in all-dielectric oligomers
,”
Nano Lett.
12
,
6459
6463
(
2012
).
28.
C.
Fang
,
Q.
Yang
,
Q.
Yuan
,
X.
Gan
,
J.
Zhao
,
Y.
Shao
,
Y.
Liu
,
G.
Han
, and
Y.
Hao
,
Opto-Electron. Adv.
4
,
200030
(
2021
).
29.
Q.
Deng
,
J.
Chen
,
L.
Long
,
B.
Chen
,
H.
Yu
, and
Z.
Li
,
Opto-Electron. Adv.
5
,
210024
(
2022
).
30.
E.
Calandrini
,
A.
Cerea
,
F.
De Angelis
,
R. P.
Zaccaria
, and
A.
Toma
,
Nanophotonics
8
,
45
62
(
2019
).
31.
D.
Smith
,
W.
Padilla
,
D.
Vier
,
S.
Nemat-Nasser
, and
S.
Schultz
,
Phys. Rev. Lett.
84
,
4184
4187
(
2000
).
32.
T. J.
Yen
,
W. J.
Padilla
,
N.
Fang
,
D. C.
Vier
,
D. R.
Smith
,
J. B.
Pendry
,
D. N.
Basov
, and
X.
Zhang
,
Science
303
,
1494
1496
(
2004
).
33.
S.
Linden
,
C.
Enkrich
,
M.
Wegener
,
J.
Zhou
,
T.
Koschny
, and
C. M.
Soukoulis
,
Science
306
,
1351
1353
(
2004
).
34.
C.
Enkrich
,
M.
Wegener
,
S.
Linden
,
S.
Burger
,
L.
Zschiedrich
,
F.
Schmidt
,
J.
Zhou
,
T.
Koschny
, and
C.
Soukoulis
,
Phys. Rev. Lett.
95
,
203901
(
2005
).
35.
P.
Wu
,
C.
Liao
,
V.
Savinov
,
T.
Chung
,
W.
Chen
,
Y.
Huang
,
P.
Wu
,
Y.
Chen
,
A.
Liu
,
N. I.
Zheludev
, and
D.
Tsai
,
ACS Nano
12
,
1920
1927
(
2018
).
36.
W.
Tsai
,
T.
Chung
,
H.
Hsiao
,
J.
Chen
, and
D.
Tsai
,
Adv. Mater.
31
,
1806479
(
2018
).
37.
W.
Yang
,
Q.
Liu
,
H.
Wang
,
Y.
Chen
,
R.
Yang
,
S.
Xia
,
Y.
Luo
,
L.
Deng
,
J.
Qin
,
H.
Duan
, and
L.
Bi
,
Nat. Commun.
13
,
1719
(
2022
).
38.
R.
Degl’Innocenti
,
D. S.
Jessop
,
Y. D.
Shah
,
J.
Sibik
,
J. A.
Zeitler
,
P. R.
Kidambi
,
S.
Hofmann
,
H. E.
Beere
, and
D. A.
Ritchie
, “
Low-bias terahertz amplitude modulator based on split-ring resonators and graphene
,”
ACS Nano
8
,
2548
2554
(
2014
).
39.
H.
Liu
,
D. A.
Genov
,
D. M.
Wu
,
Y. M.
Liu
,
J. M.
Steele
,
C.
Sun
,
S. N.
Zhu
, and
X.
Zhang
,
Phys. Rev. Lett.
97
,
243902
(
2006
).
40.
N.
Liu
,
H.
Guo
,
L.
Fu
,
S.
Kaiser
,
H.
Schweizer
, and
H.
Giessen
,
Nat. Mater.
7
,
31
37
(
2008
).
41.
N.
Liu
,
S.
Kaiser
, and
H.
Giessen
,
Adv. Mater.
20
,
4521
4525
(
2008
).
42.
I.
Sersic
,
M.
Frimmer
,
E.
Verhagen
, and
A. F.
Koenderink
,
Phys. Rev. Lett.
103
,
213902
(
2009
).
43.
N.
Feth
,
K.
Michael
,
M.
Husnik
,
S.
Kai
, and
S.
Linden
,
Opt. Express
18
,
6545
6554
(
2010
).
44.
H.
Liu
,
D. A.
Genov
,
D.
Wu
,
Y.
Liu
,
Z.
Liu
,
C.
Sun
,
S.
Zhu
, and
X.
Zhang
,
Phys. Rev. B
76
,
073101
(
2007
).
45.
T.
Li
,
H.
Liu
,
T.
Li
,
S.
Wang
,
J.
Cao
,
Z.
Zhu
,
Z.
Dong
,
S.
Zhu
, and
X.
Zhang
,
Phys. Rev. B
80
,
115113
(
2009
).
46.
T.
Li
,
R.
Ye
,
C.
Li
,
H.
Liu
,
S.
Wang
,
J.
Cao
,
S.
Zhu
, and
X.
Zhang
,
Opt. Express
17
,
11486
11494
(
2009
).
47.
J.
Chen
,
W.
Fan
,
T.
Zhang
,
C.
Tang
,
X.
Chen
,
J.
Wu
,
D.
Li
, and
Y.
Yu
,
Opt. Express
25
,
3675
3681
(
2017
).
48.
H.
Liu
,
T.
Li
,
Q. J.
Wang
,
Z. H.
Zhu
,
S. M.
Wang
,
J. Q.
Li
,
S. N.
Zhu
,
Y. Y.
Zhu
, and
X.
Zhang
, “
Extraordinary optical transmission induced by excitation of a magnetic plasmon propagation mode in a diatomic chain of slit-hole resonators
,”
Phys. Rev. B
79
,
024304
(
2009
).
49.
J.
Wang
,
C.
Fan
,
J.
He
,
P.
Ding
,
E.
Liang
, and
Q.
Xue
, “
Double fano resonances due to interplay of electric and magnetic plasmon modes in planar plasmonic structure with high sensing sensitivity
,”
Opt. Express
21
,
2236
2244
(
2013
).
50.
Z.
Liu
,
S.
Du
,
A.
Cui
,
Z.
Li
,
Y.
Fan
,
S.
Chen
,
W.
Li
,
J.
Li
, and
C.
Gu
, “
High-quality-factor mid-infrared toroidal excitation in folded 3D metamaterials
,”
Adv. Mater.
29
,
1606298
(
2017
).
51.
Y.
Imade
,
R.
Ulbricht
,
M.
Tomoda
,
O.
Matsuda
,
G.
Seniutinas
,
S.
Juodkazis
, and
O. B.
Wright
, “
Gigahertz optomechanical modulation by split-ring-resonator nanophotonic meta-atom arrays
,”
Nano Lett.
17
,
6684
6689
(
2017
).
52.
P.
Wu
,
T.
Wei
,
K.
Yang
,
C.
Hsiao
,
G.
Sun
,
Q.
Ai
,
N.
Zheludev
, and
D.
Tsai
,
Nanophotonics
1
,
131
138
(
2012
).
53.
J.
Chen
,
C.
Yang
,
P.
Gu
,
Y.
Kuang
, and
Z.
Liu
,
J. Lightwave Technol.
39
,
562
565
(
2021
).
54.
Z.
Yan
,
X.
Lu
,
K.
Chen
,
Z.
Lv
,
X.
Pu
,
C.
Tang
, and
P.
Cai
,
J. Lightwave Technol.
39
,
7217
7222
(
2021
).
55.
C.
Jing
,
N.
Hai
,
P.
Cheng
,
S.
Qi
,
C.
Tang
,
Z.
Ying
,
L.
Wang
, and
P.
Gun-Sik
,
J. Lightwave Technol.
36
,
3481
3485
(
2018
).
56.
J.
Chen
,
P.
Mao
,
R.
Xu
,
C.
Tang
,
Y.
Liu
,
Q.
Wang
, and
L.
Zhang
,
Opt. Express
23
,
16238
16245
(
2015
).
57.
N.
Liu
,
H.
Liu
,
S.
Zhu
, and
H.
Giessen
,
Nat. Photonics
3
,
157
162
(
2009
).
58.
Z.
Wang
,
Y.
Hu
,
X.
Xiong
,
R.
Peng
, and
M.
Wang
,
Opt. Lett.
42
,
1153
1156
(
2017
).
59.
P.
Wu
,
G.
Sun
,
W.
Chen
,
K.
Yang
,
Y.
Huang
,
Y.
Chen
,
H.
Huang
,
W.
Hsu
,
H.
Chiang
, and
D.
Tsai
,
Appl. Phys. Lett.
105
,
33105
(
2014
).
60.
W.
Hsu
,
P.
Wu
,
J.
Chen
,
T.
Chen
, and
D.
Tsai
,
Sci. Rep.
5
,
11226
(
2015
).
61.
F. B. P.
Niesler
,
S.
Linden
,
J.
Forstner
,
Y.
Grynko
,
T.
Meier
, and
M.
Wegener
,
Phys. Rev. Lett.
109
,
015502
(
2012
).
62.
K.
O’Brien
,
H.
Suchowski
,
J.
Rho
,
A.
Salandrino
,
B.
Kante
,
X.
Yin
, and
X.
Zhang
,
Nat. Mater.
14
,
379
383
(
2015
).
63.
V. M.
Shalaev
,
W.
Cai
,
U.
Chettiar
,
H. K.
Yuan
, and
A. V.
Kildishev
,
Opt. Lett.
30
,
3356
3358
(
2005
).
64.
S.
Zhang
,
W.
Fan
,
K. J.
Malloy
,
S. R. J.
Brueck
, and
R. M.
Osgood
,
Opt. Express
13
,
4922
4930
(
2005
).
65.
G.
Dolling
,
C.
Enkrich
,
M.
Wegener
,
C. M.
Soukoulis
, and
S.
Linden
,
Opt. Lett.
31
,
1800
1802
(
2006
).
66.
S.
Zhang
,
W.
Fan
,
K. J.
Malloy
,
S. R. J.
Brueck
,
N. C.
Panoiu
, and
R. O.
Osgood
,
J. Opt. Soc. Am. B
23
,
434
438
(
2006
).
67.
S.
Zhang
,
W.
Fan
,
N. C.
Panoiu
,
K. J.
Malloy
,
R. M.
Osgood
, and
S. R.
Brueck
,
Opt. Express
14
,
6778
(
2006
).
68.
G.
Dolling
,
M.
Wegener
, and
S.
Linden
,
Opt. Lett.
32
,
551
553
(
2007
).
69.
G.
Dolling
,
M.
Wegener
,
C. M.
Soukoulis
, and
S.
Linden
,
Opt. Lett.
32
,
53
55
(
2007
).
70.
C.
García-Meca
,
R.
Ortuo
,
F. J.
Rodríguez-Fortuo
,
J.
Martí
, and
A.
Martínez
,
Opt. Lett.
34
,
1603
1605
(
2009
).
71.
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
).
72.
J.
Valentine
,
S.
Zhang
,
T.
Zentgraf
,
E.
Ulin-Avila
,
D. A.
Genov
,
G.
Bartal
, and
X.
Zhang
,
Nature
455
,
376
(
2008
).
73.
M. C.
Garcia
,
J.
Hurtado
,
J.
Marti
,
A.
Martinez
,
W.
Dickson
, and
A. V.
Zayats
,
Phys. Rev. Lett.
106
,
067402
(
2011
).
74.
S.
Xiao
,
V. P.
Drachev
,
A. V.
Kildishev
,
X.
Ni
,
U. K.
Chettiar
,
H. K.
Yuan
, and
V. M.
Shalaev
,
Nature
466
,
735–U6
(
2010
).
75.
L.
Wang
,
A. S.
Shorokhov
,
P. N.
Melentiev
,
S.
Kruk
,
M.
Decker
,
C.
Helgert
,
F.
Setzpfandt
,
A. A.
Fedyanin
,
Y. S.
Kivshar
, and
D. N.
Neshev
,
ACS Photonics
3
,
1494
1499
(
2016
).
76.
V. A.
Podolskiy
,
A. K.
Sarychev
, and
V. M.
Shalaev
,
Opt. Express
11
,
735
745
(
2003
).
77.
L. V.
Panina
,
A. N.
Grigorenko
, and
D. P.
Makhnovskiy
,
Phys. Rev. B
66
,
1155411
(
2002
).
78.
A. N.
Lagarkov
and
A. K.
Sarychev
,
Phys. Rev. B
53
,
6318
6336
(
1996
).
79.
V. A.
Podolskiy
,
A. K.
Sarychev
, and
V. M.
Shalaev
,
J. Nonlin. Opt. Phys. Mater.
11
(1),
65
74
(
2002
).
80.
J. B.
Pendry
,
D.
Schurig
, and
D. R.
Smith
, “
Controlling electromagnetic fields
,”
Science
312
,
1780
1782
(
2006
).
81.
J. B.
Pendry
, “
Negative refraction makes a perfect lens
,”
Phys. Rev. Lett.
85
,
3966
3969
(
2000
).
82.
N.
Fang
,
H.
Lee
,
C.
Sun
, and
X.
Zhang
, “
Sub-diffraction-limited optical imaging with a silver superlens
,”
Science
308
,
534
537
(
2005
).
83.
N.
Engheta
,
A.
Salandrino
, and
A.
Alu
,
Phys. Rev. Lett.
95
,
095504
(
2005
).
84.
T.
Li
,
H.
Liu
,
F.
Wang
,
Z.
Dong
, and
X.
Zhang
,
Opt. Express
14
,
11155
11163
(
2006
).
85.
M. A.
Noginov
,
G.
Zhu
,
M.
Bahoura
,
J.
Adegoke
,
C. E.
Small
,
B. A.
Ritzo
,
V. P.
Drachev
, and
V. M.
Shalaev
,
Opt. Lett.
31
,
3022
3024
(
2006
).
86.
S. A.
Ramakrishna
and
J. B.
Pendry
,
Phys. Rev. B
67
,
201101
(
2003
).
87.
A. K.
Sarychev
and
G.
Tartakovsky
,
Phys. Rev. B
75
,
085436
(
2006
).
88.
A.
Fang
,
T.
Koschny
,
M.
Wegener
, and
C. M.
Soukoulis
,
Phys. Rev. B
79
,
241104
(
2009
).
89.
Y.
Sivan
,
S.
Xiao
,
U. K.
Chettiar
,
A. V.
Kildishev
, and
V. M.
Shalaev
,
Opt. Express
17
,
24060
24074
(
2009
).
90.
K. M.
Dani
,
Z. Y.
Ku
,
P. C.
Upadhya
,
R. P.
Prasankumar
,
S. R. J.
Brueck
, and
A. J.
Taylor
,
Nano Lett.
9
,
3565
3569
(
2009
).
91.
A.
Minovich
,
J.
Farnell
,
D. N.
Neshev
,
I.
Mckerracher
,
F.
Karouta
,
J.
Tian
,
D. A.
Powell
,
I. V.
Shadrivov
,
H. H.
Tan
,
C.
Jagadish
, and
Y. S.
Kivshar
,
Appl. Phys. Lett.
100
,
125
(
2012
).
92.
J.
Reinhold
,
M. R.
Shcherbakov
,
A.
Chipouline
,
V. I.
Panov
,
C.
Helgert
,
T.
Paul
,
C.
Rockstuhl
,
F.
Lederer
,
E.-B.
Kley
,
A.
Tuennermann
,
A. A.
Fedyanin
, and
T.
Pertsch
,
Phys. Rev. B
86
,
115401
(
2012
).
93.
L. P.
Wang
and
Z. M.
Zhang
,
Appl. Phys. Lett.
100
,
063902
(
2012
).
94.
L. S.
Long
,
S.
Taylor
, and
L. P.
Wang
,
ACS Photonics
7
,
2219
2227
(
2020
).
95.
A.
Ishikawa
,
S.
Zhang
,
D. A.
Genov
,
G.
Bartal
, and
X.
Zhang
,
Phys. Rev. Lett.
102
,
043904
(
2009
).
96.
J. A.
Dionne
,
L. A.
Sweatlock
,
H. A.
Atwater
, and
A.
Polman
,
Phys. Rev. B
73
,
035407
(
2006
).
97.
N. J.
Gollub
,
D. R.
Smith
,
D. C.
Vier
,
T.
Perram
, and
J. J.
Mock
,
Phys. Rev. B
71
,
195402
(
2005
).
98.
R.
Verre
,
Z. J.
Yang
,
T.
Shegai
, and
M.
Käll
,
Nano Lett.
15
,
1952
1958
(
2015
).
99.
Y.
Jin
,
J.
Liang
,
S.
Wu
,
Y.
Zhang
,
L.
Zhou
,
Q.
Wang
,
H.
Liu
, and
J.
Zhu
,
Adv. Mater.
32
,
2000058
(
2020
).
100.
M. H.
Ryou
,
Y. M.
Lee
,
Y.
Lee
,
M.
Winter
, and
P.
Bieker
,
Adv. Funct. Mater.
25
,
834
841
(
2015
).
101.
A.
Moreau
,
C.
Ciracì
,
J. J.
Mock
,
R. T.
Hill
,
Q.
Wang
,
B. J.
Wiley
,
A.
Chilkoti
, and
D. R.
Smith
,
Nature
492
,
86
89
(
2012
).
102.
N.
Liu
,
M.
Mesch
,
T.
Weiss
,
M.
Hentschel
, and
H.
Giessen
,
Nano Lett.
10
,
2342
2348
(
2010
).
103.
S.
Chen
,
Y.
Zhang
,
T. M.
Shih
,
W.
Yang
,
S.
Hu
,
X.
Hu
,
J.
Li
,
B.
Ren
,
B.
Mao
,
Z.
Yang
, and
Z.
Tian
,
Nano Lett.
18
,
2209
2216
(
2018
).
104.
Y.
Meng
,
Q.
Zhang
,
D.
Lei
,
Y.
Li
,
S.
Li
,
Z.
Liu
,
W.
Xie
, and
C. W.
Leung
,
Laser Photonics Rev.
14
,
2000068
(
2020
).
105.
L.
Zhao
,
Z.
He
,
Q.
Niu
,
X.
Yang
, and
S.
Dong
,
J. Quant. Spectrosc. Radiat. Therapy
240
,
106631
(
2014
).
106.
H.
Yoo
,
J.
Shin
,
J.
Sim
,
H.
Cho
, and
S.
Hong
,
Biosens. Bioelectron.
168
,
112561
(
2020
).
107.
V. A.
Fedotov
,
P. L.
Mladyonov
,
S. L.
Prosvirnin
, and
N. I.
Zheludev
,
Phys. Rev. E
72
,
056613
(
2005
).
108.
W.
Cai
,
U. K.
Chettiar
,
H. K.
Yuan
,
V. C.
de Silva
,
A. V.
Kildishev
,
V. P.
Drachev
, and
V. M.
Shalaev
,
Opt. Express
15
,
3333
3341
(
2007
).
109.
S.
Mubeen
,
S.
Zhang
,
N.
Kim
,
S.
Lee
,
S.
Krämer
,
H.
Xu
, and
M.
Moskovits
,
Nano Lett.
12
,
2088
2094
(
2012
).
110.
P.
Nordlander
and
E.
Prodan
,
Nano Lett.
4
,
2209
2213
(
2004
).
111.
J. J.
Mock
,
R. T.
Hill
,
A.
Degiron
,
S.
Zauscher
,
A.
Chilkoti
, and
D. R.
Smith
,
Nano Lett.
8
,
2245
2252
(
2008
).
112.
P.
Vasa
,
W.
Wang
,
R.
Pomraenke
,
M.
Lammers
,
M.
Maiuri
,
C.
Manzoni
,
G.
Cerullo
, and
C.
Lienau
,
Nat. Photonics
7
,
128
132
(
2013
).
113.
S.
Raza
,
N.
Stenger
,
A.
Pors
,
T.
Holmgaard
,
S.
Kadkhodazadeh
,
J. B.
Wagner
,
K.
Pedersen
,
M.
Wubs
,
S. I.
Bozhevolnyi
, and
N. A.
Mortensen
,
Nat. Commun.
5
,
4125
(
2014
).
114.
W.
Wang
,
P.
Vasa
,
R.
Pomraenke
,
R.
Vogelgesang
,
S. A.
De
,
E.
Sommer
,
M.
Maiuri
,
C.
Manzoni
,
G.
Cerullo
, and
C.
Lienau
,
ACS Nano
8
,
1056
1064
(
2014
).
115.
W.
Wang
,
P.
Vasa
,
E.
Sommer
,
S. A.
De
,
P.
Gross
,
R.
Vogelgesang
, and
C.
Lienau
,
J. Optics-UK
16
,
114021
(
2014
).
116.
W.
Wang
,
H.
Zhang
,
R.
Vogelgesang
,
P.
Vasa
, and
C.
Lienau
,
Appl. Phys. Lett.
110
,
151101
(
2017
).
117.
D.
Li
,
J.
Ou
,
P.
Xie
,
Z.
Liang
,
W.
Wang
,
W.
Wang
,
H.
Zhang
, and
X.
Kuang
,
Phys. Rev. B
103
,
245404
(
2021
).
118.
Y.
Shen
,
K.
He
,
Q. S.
Zou
,
S.
Xing
,
J. Y.
Huang
,
M. C.
Zheng
,
X. Y.
She
, and
C. J.
Jin
,
Adv. Funct. Mater.
32
,
2108741
(
2022
).
119.
Y.
Zhang
,
Q.
Jiang
,
M.
Long
,
R.
Han
,
K.
Cao
,
S.
Zhang
,
D.
Feng
,
T.
Jia
,
Z.
Sun
,
J.
Qiu
, and
H.
Xu
,
Opto-Electron. Sci.
1
,
220005
(
2022
).
120.
B.
Zhao
,
J. M.
Zhao
, and
Z. M.
Zhang
,
Appl. Phys. Lett.
105
,
031905
(
2014
).
121.
W.
Wang
,
Y.
Zhu
,
H.
Zhang
,
L.
Qing
,
L.
Lei
,
D.
Li
, and
J.
Yi
,
Phys. Rev. B
99
,
235407
(
2019
).
122.
D. M.
Li
,
X. Y.
Kuang
,
H.
Zhang
,
Y. Z.
Liang
,
T.
Xu
,
L. Y.
Qing
,
Y. H.
Zhu
,
S.
Zhang
,
W. X.
Wang
, and
W.
Wang
,
Opt. Express
26
,
34122
34130
(
2018
).
123.
L. P.
Wang
and
Z. M.
Zhang
,
Appl. Phys. Lett.
95
,
111904
(
2009
).
124.
B.
Zhao
and
Z. M.
Zhang
,
J. Quant. Spectrosc. Radiat. Therapy
135
,
81
89
(
2014
).
125.
Y.
Zhu
,
H.
Zhang
,
D.
Li
,
Z.
Zhang
,
S.
Zhang
,
J.
Yi
, and
W.
Wang
,
Opt. Express
26
,
9148
9154
(
2018
).
126.
Q.
Pan
,
G.
Zhang
,
R.
Pan
,
J.
Zhang
,
Y.
Shuai
, and
H.
Tan
,
Opt. Express
26
,
18230
18237
(
2018
).
127.
Y. M.
Qing
,
H. F.
Ma
, and
T. J.
Cui
,
ACS Photonics
6
,
2884
2893
(
2019
).
128.
C.
Zeng
,
H.
Lu
,
D.
Mao
,
Y.
Du
,
H.
Hua
,
W.
Zhao
, and
J.
Zhao
,
Opto-Electron Adv.
5
,
200098
(
2022
).
129.
C.
Strobel
,
C. A.
Chavarin
,
B.
Leszczynska
,
S.
Leszczynski
,
F.
Winkler
,
S.
Killge
,
S.
Völkel
,
K.
Richter
,
A.
Hiess
,
M.
Knaut
,
J.
Reif
,
M.
Albert
,
C.
Wenger
, and
J. W.
Bartha
,
J. Appl. Phys.
125
,
234501
(
2019
).
130.
C.
Yan
,
J. H.
Cho
, and
J. H.
Ahn
,
Nanoscale
4
,
4870
4882
(
2012
).
131.
M.
Kim
,
P.
Kang
,
J.
Leem
, and
S.
Nam
,
Nanoscale
9
,
4058
4065
(
2017
).
132.
X.
Liu
,
E. K.
Lee
, and
J. H.
Oh
,
Small
10
,
3700
3706
(
2014
).
133.
S.
Yan
,
Y.
Zuo
,
S.
Xiao
,
L. K.
Oxenløwe
, and
Y.
Ding
,
Opto-Electron. Adv.
5
,
210159
(
2022
).
134.
Y.
Yao
,
M. A.
Kats
,
R.
Shankar
,
Y.
Song
,
J.
Kong
,
M.
Loncar
, and
F.
Capasso
,
Nano Lett.
14
,
214
219
(
2014
).
135.
X.
Miao
,
S.
Tongay
,
M. K.
Petterson
,
K.
Berke
,
A. G.
Rinzler
,
B. R.
Appleton
, and
A. F.
Hebard
,
Nano Lett.
12
,
2745
2750
(
2012
).
136.
C.
Zhu
,
H.
Liu
,
S. M.
Wang
,
T.
Li
,
J. X.
Cao
,
Y. J.
Zheng
,
L.
Li
,
Y.
Wang
,
S. N.
Zhu
, and
X.
Zhang
,
Opt. Express
18
,
26268
26273
(
2010
).
137.
C.
Cherqui
,
Y.
Wu
,
G.
Li
,
S. C.
Quillin
,
J. A.
Busche
,
N.
Thakkar
,
C. A.
West
,
N. P.
Montoni
,
P. D.
Rack
, and
J. P.
Camden
,
Nano Lett.
16
,
6668
6676
(
2016
).
138.
N. J.
Greybush
,
P. V.
Pacheco
,
N.
Engheta
,
C. B.
Murray
, and
C. R.
Kagan
,
ACS Nano
13
,
1617
1624
(
2019
).
139.
N. P.
Montoni
,
S. C.
Quillin
,
C.
Cherqui
, and
D. J.
Masiello
,
ACS Photonics
5
,
3272
3281
(
2018
).
140.
G.
Li
,
Q.
Li
, and
L.
Wu
,
Nanoscale
7
,
19914
19920
(
2015
).
141.
Y.
Bao
,
Z.
Hu
,
Z.
Li
,
X.
Zhu
, and
Z.
Fang
,
Small
11
,
2177
2181
(
2015
).
142.
W. L.
Barnes
,
A.
Dereux
, and
T. W.
Ebbesen
,
Nature
424
,
824
830
(
2003
).
143.
M. L.
Brongersma
,
J. W.
Hartman
, and
H. A.
Atwater
,
Phys. Rev. B
62
,
16356
16359
(
2000
).
144.
D.
Dregely
,
M.
Hentschel
, and
H.
Giessen
,
ACS Nano
5
,
8202
8211
(
2011
).
145.
N.
Engtieta
,
Science
317
,
1698
1702
(
2007
).
146.
J. A.
Fan
,
C.
Wu
,
K.
Bao
,
J.
Bao
,
R.
Bardhan
,
N. J.
Halas
,
V. N.
Manoharan
,
P.
Nordlander
,
G.
Shvets
, and
F.
Capasso
,
Science
328
,
1135
1138
(
2010
).
147.
M.
Hentschel
,
D.
Dregely
,
R.
Vogelgesang
,
H.
Giessen
, and
N.
Liu
,
ACS Nano
5
,
2042
2050
(
2011
).
148.
F.
Shafiei
,
F.
Monticone
,
K. Q.
Le
,
X.
Liu
,
T.
Hartsfield
,
A.
Alu
, and
X.
Li
,
Nat. Nanotechnol.
8
,
95
99
(
2013
).
149.
N.
Liu
,
S.
Mukherjee
,
K.
Bao
,
L. V.
Brown
,
J.
Dorfmueller
,
P.
Nordlander
, and
N. J.
Halas
,
Nano Lett.
12
,
364
369
(
2012
).
150.
D.
Yang
,
S.
Im
,
G.
Pan
,
S.
Ding
,
Z.
Yang
,
Z.
Hao
,
L.
Zhou
, and
Q.
Wang
,
Nanoscale
9
,
6068
6075
(
2017
).
151.
N. S.
King
,
M. W.
Knight
,
N.
Large
,
A. M.
Goodman
, and
N. J.
Halas
,
Nano Lett.
13
,
5997
6001
(
2013
).
152.
S.
Ding
,
H.
Zhang
,
D.
Yang
,
Y.
Qiu
,
F.
Nan
,
Z.
Yang
,
J.
Wang
,
Q.
Wang
, and
H.
Lin
,
Nano Lett.
19
,
2005
2011
(
2019
).
153.
A.
Nazir
,
S.
Panaro
,
R. P.
Zaccaria
,
C.
Liberale
,
A. F.
De
, and
A.
Toma
,
Nano Lett.
14
,
3166
3171
(
2014
).
154.
S.
Linden
,
M.
Decker
, and
M.
Wegener
,
Phys. Rev. Lett.
97
,
083902
(
2006
).
155.
R.
Ameling
and
H.
Giessen
,
Nano Lett.
10
,
4394
4398
(
2010
).
156.
C.
Zhang
,
J.
Fang
,
W.
Yang
,
Q.
Song
, and
S.
Xiao
,
Adv. Opt. Mater.
5
,
1700469
(
2017
).
157.
P.
Xie
,
Z.
Liang
,
Z.
Li
,
W.
Wang
,
W.
Wang
,
T.
Xu
,
X.
Kuang
,
L.
Qing
,
D.
Li
, and
J.
Yi
,
Phys. Rev. B
101
,
045403
(
2020
).
158.
H.
Lu
,
Z.
Yue
,
Y.
Li
,
Y.
Zhang
,
M.
Zhang
,
W.
Zeng
,
X.
Gan
,
D.
Mao
,
F.
Xiao
,
T.
Mei
,
W.
Zhao
,
X.
Wang
,
M.
Gu
, and
J.
Zhao
,
Light Sci. Appl.
9
,
191
(
2020
).
159.
P.
Xie
,
D.
Li
,
Y.
Chen
,
P.
Chang
,
H.
Zhang
,
J.
Yi
, and
W.
Wang
,
Phys. Rev. B
102
,
115430
(
2020
).
160.
M. Z.
Hasan
and
C. L.
Kane
,
Rev. Mod. Phys.
82
,
3045
3067
(
2010
).
161.
Z.
Yue
,
G.
Xue
,
J.
Liu
,
Y.
Wang
, and
M.
Gu
,
Rev. Mod. Phys.
82
,
3045
3067
(
2010
).
162.
K.
Ding
and
C. Z.
Ning
,
Light Sci. Appl.
1
,
e20
(
2012
).
163.
M.
Ren
,
M.
Chen
,
W.
Wu
,
L.
Zhang
,
J.
Liu
,
B.
Pi
,
X.
Zhang
,
Q.
Li
,
S.
Fan
, and
J.
Xu
,
Nano Lett.
15
,
2951
2957
(
2015
).