The recent development in terahertz (THz) technology enables a wide range of applications in electronics , , , , ,  and photonics , , ; for medical , , , military , , , and security purposes , , ; in microdisplay ; and even in the investigation and conservation of cultural heritages , , , which is based on its excellent sensitivity and selectivity. Such advances in THz applications have been stimulated by the improvement in functional THz devices such as filters , , , , , , switches , , , , , , and sensors , , , , , . Artificially structured metamaterials in subwavelength scale, especially, provide us with the ability to control THz responses over the broad bandwidths , , , , , , determined by key parameters: length scale , , geometry , and surrounding materials , . The dielectric properties of these metamaterials, changeable by optical , , , , electrical , , thermal , , , or other mechanical , ,  means, with or without lateral patterns, are in turn used to control transmission or reflection behavior over entire THz frequencies, further extending device performance. THz wave interaction with deep subwavelength nanostructures with extremely high cross-sections will be discussed in this review as introduced with several milestones regarding the structures from micrometer , , , , , ,  to nanometer , , , , , , , , , ,  scale and, finally, the Angstrom scales ,  (Figure 1). In particular, the enhanced THz nonlinear phenomena in Angstrom-sized infinite gaps will find unprecedented resonant and nonresonant applications in metamaterials , , , , , . Finally, several novel types of THz sensing technology based on the field enhancement properties using subwavelength structures for variety of chemistry, biology, and medical applications will be introduced.
2 Fundamentals of giant terahertz field enhancement at an infinitely long nanogap
Using the vector Babinet principle, concentrating terahertz (THz) electric field in a small aperture in metal , , , , , ,  is equivalent to enhancing the magnetic field of light around rod and patch antennas , , , , , , , . Although a metal film in THz frequency is neither perfect conductor nor infinitely flat, various areas of THz research , , , , , , , , , , , , , , , , , including THz plasmonics , , , , , , ,  and metamaterials , , , , , , , implicitly use this principle, which remains true in its spirit if not in its rigor. One of the earliest theoretical accounts on the light-aperture interaction had been made by Sommerfeld’s half-plane problem , ,  that was revisited by Bethe and Bouwkamp to deal with small apertures in infinitely thin perfect conducting plates , , , . These earlier studies instigated further investigations on the light interaction with periodic aperture systems , which were later extended to studies on enhanced light transmission through the apertures in visible , , , infrared (IR) , and THz , , , ,  regimes. To resolve the enhanced transmission and consequent strong field enhancement near the apertures, numerous theoretical schemes such as coupled-mode theory , , , , , , , , , microscopic models , , , transfer matrix method , , and capacitor model , incorporated with various numerical methods , , , , were developed.
Basically, the field enhancement can be explained in terms of macroscopic accumulation of surface charges near the apertures’ edges, driven by incident light. The accumulated charges give rise to capacitive enhancement of electric field in the gap , . This simple picture paved a way to the gigantic field enhancement supported by deep subwavelength apertures in metal (Figure 2), which is, in particular, essential for THz nonlinear optics , , , , , ,  that requires intense electromagnetic field.
In this section, we discuss THz field enhancement in deep subwavelength apertures. Specifically, we focus on nonresonant THz field enhancement by an infinitely long gap with sub-skin-depth thickness and provide a simple model to explain that the field enhancement can be simply given by a ratio of photon wavelength and thickness. We also discuss a Kirchhoff integral formalism that allows quantitative estimation of field enhancement in the near-field from the far-field measurement. After that, we discuss resonant field enhancement supported by a rectangle slot, exhibiting different behavior of field enhancement compared to the non-resonant case.
2.1 THz field enhancement and skin-depth physics
We first concentrate on metal films of thickness 100 nm or less, satisfying the sub-skin-depth condition at THz regime . Consider, for a good conductor, that the ultimate field enhancement (FE) in a high-aspect ratio nanogap of air sandwiched between two sub-skin-depth metallic planes, irradiated by an electromagnetic wave with an incident electric field of Einc, has a simple analytical expression when the gap width w is much smaller than the film thickness h:
where Egap is the electric field at the gap, λ is the vacuum wavelength, and h the film thickness , , . When the gap is filled with a dielectric of permittivity, εgap, the formula is modified to, somewhat trivially,
where ε0 is the vacuum permittivity . What is surprising is that the gap FE is seemingly independent of how good the metal is.
To derive Eq. (1), we first consider direct transmission of electromagnetic wave through a thin film of free-standing metal; for the sake of simplicity, we assume a normal incidence. For this instance, by a thin film. We mean films whose thickness h is smaller than the skin depth but larger than a characteristic thickness of
where σm is the conductivity of metal, μ0 is the vacuum magnetic permeability, and ω=2πf is the angular frequency of the electromagnetic wave. The characteristic thickness, h0, at which absorption loss by the metal is 50% is only 0.53 nm for a reasonably good metal of a conductivity 107 Siemens per meter (Ω−1m−1), so that any transition metal films of today’s technology fall above this thickness. Skin depth at 1 THz is typically 100 nm or more even for good metals so that transition metal films in the range of 5–100 nm thickness satisfy Eq. (3) (Figure 3). For these thin metal films, the amplitude transmission and reflection coefficients, t and r, respectively, are given as
where Einc is the incident electric field and Et is the transmitted electric field as described in Figure 4.
To derive Eq. (4), we consider transverse magnetic polarized incident light (Ex, Hy) and focus on the magnetic field of light near the thin film. At the incident surface, reflection makes the magnetic field of light approximately twice that of the incident field, whereas on the transmission side, the magnetic field is much smaller than the incident field. We assume a constant electric field/current density inside and apply Ampere’s law, ∇×H=J, which gives rise to the current density ignoring the vacuum displacement-current term. Then, the continuity of tangential component of electric fields at the air-metal interface leads to resulting in Eq. (4) (Figure 5A). Here, Em is electric field just inside the metal surface of the transmitting side. Since the normal component of the displacement current is the same across the air-metal boundary of the gap, we obtain having taken advantage of the metal dielectric constant in the terahertz regime
where ε∞ is the high-frequency dielectric constant of the metal and γ is the Drude damping constant (Figure 5B). The physics of the conductivity independence of the FE is then clear: better conductivity makes the field inside the metal weaker, which is compensated by the higher dielectric constant of the metal when applying the displacement current boundary condition. Equation (1) can be analytically extended into a completely different regime, for samples much thicker than the skin depth (Figure 5C); for this regime, we assume that current flows only at the surface with surface current density K=2Hinc, and terminates at the exit side of the gap, dumping charges along the way, as shown in Figure 6. Interestingly, Jx component at the metal surface in the gap, shown in Figure 6C, is kept nearly constant, while Jz (Figure 6B) is not. This implies that the accumulated surface charge is evenly distributed over the metal surface, giving rise to constant Ex field in the gap. The continuity equation demands the surface charge density being from which follows.
Aforementioned FE with thin (w~h<δ) and thick (w<δ<h) narrow gaps, together with wide gaps (δ, h<w), are discussed in a recent paper . Shown in Figure 7A are numerically calculated current distributions with narrow (1.5 nm) and wide (200 nm) gaps in 150-nm-thick gold films, and corresponding schematics for charge distributions are shown in Figure 7B. For the wider gap, charges are not mostly accumulated at the metal edges in the gap but are spread over the surface outside the gap. This consequently reduces the FE. For the narrower gap, most charges are accumulated at the metal edge in the gap, giving rise to stronger FE. However, we stress that when the aspect ratio w/h is sufficiently small (w/h≪1), the charge distribution becomes insensitive to the gap size, so that the FE will exhibit saturating behavior with decreasing w. Also, 1/h dependence of FE in narrow gap is well demonstrated (Figure 7C).
Having established a simple FE formulae for thin and thick samples limits, we proceed to include the very thinregime, where t can no longer assumed to be ≪1. Applying Ampere’s law, we obtain Applying energy conservation including Ohmic loss, we arrive at Solving these two, we get
with h0 given by Eq. (3); Eq. (6) can now be applied to all three regimes of thicknesses.
To confirm that Eq. (6) gives a reasonably good description of the FE over many orders of magnitudes of conductivity, we plot, in Figure 8A, finite-difference-time-domain (FDTD) calculations (solid red line) for a 50-nm-thick film with 1 nm air gap varying conductivity over five orders of magnitudes. We also plot the FE of Eq. (6) (dash-dot line), showing good agreements. While the metal film in Figure 8A belongs to a very thin to thin regime, the h=1000 nm case shown in Figure 8B certainly belongs to the thick regime for σm≥106 Ω−1m−1; skin depth at 1 THz is 500 nm for a conductivity of σm=106 Ω−1m−1.
Next, we consider the effects of a substrate. Assuming an infinitely thick substrate and applying energy conservation and Ampere’s law, it is straightforward to show that
where ns is the index of refraction of the substrate and the electromagnetic field is incident from the air side. In obtaining Eq. (7), as before, it has been assumed that the current flows straight into the gap. However, even for the high aspect ratio gaps, the asymmetry caused by the substrate would bend the current flow, resulting in loss of FE.
Although the result of FE from our model is surprisingly simple, we would like to emphasize that the same FE rule can be obtained from more rigorous analytic theory. Coupled mode theory with single-mode approximation, dealing with boundary conditions of Maxwell’s theory with single quantized waveguide mode in the gap, provides that the FE by a gap in very good conductor on the substrate can be written as 
where W0 and WS are the light-gap coupling factors defined by
with being the Hankel function of the first kind. For a narrow gap (w≪λ), the coupling factors can be approximated as
where n0=1 and γE is the Euler constant. The explicit form of FE is somewhat complicated, but clearly, a further simplification can be made by the fact that the coupling factor approaches zero as the gap size approaches a deep subwavelength regime. Interestingly, one can readily find from Eqs. (8) and (10) that, in the limit of w≪h≪, regardless of the substrate, Eq. (8) can be reduced to
which perfectly coincides with our main result of FE in Eqs. (1) and (7) with h0≪h. Even though the FE in the ultimate limit is the same with the free-standing case, however, we point out that the FE with a substrate will be generally lower. This is related to the nature of logarithmic dependency of the coupling factor in Eq. (10), which yields, in the presence of the substrate, slower vanishing of the coupling factor as w approaches zero. We point out that the physical account of the reduced FE can be found from the impedance mismatch at the air-substrate interface in the gap that naturally leads to reduction of the tangential component of the electric field near the surface .
It should be also noted that, for periodic arrays of slit, FE is limited by the periodicity especially when the period is in subwavelength scale, as shown in Figure 9 . This is an intuitive result in that the enhancement is basically related to the funneling of electromagnetic energy and that periodic slits share available energy from the incident wave, giving rise to reduced FE in each slit.
Another relevant physics that should be addressed is the enhancement of the magnetic field with metal nanowire, a Babinet complementary structure of nanogap (shown in Figure 10). Although the strict Babinet principle is applicable only to structures of an infinitesimally thin perfect electric conductor, Koo et al.  theoretically demonstrated that a qualitative prediction of FE in a complementary structure can be doable from a metal of finite thickness and conductivity.
2.2 Estimation of the field enhancement from far-field measurements: Kirchhoff integral method
Experimental verification of the huge FE by the direct measurement of near-field is a technically challenging issue because of the limited resolution of the near-field imaging, restricted by the spot size of optical probe beam, which is usually in a few micrometers , , , , , , , . To improve this issue, a more elaborately complemented scheme is introduced, of which a main idea is based on the Kirchhoff diffraction theory ,  that provides approximated relationship between the near-field information and the signal in sufficiently far distance , . A typical experimental setup utilizing the Kirchhoff method to extract the THz FE in the gap is illustrated in Figure 11. This utilizes the transmitted signal through the “normalizing aperture” as reference signal.
The scheme starts from the Kirchhoff integral expression of far-field electric field diffracted by an aperture,
where r is the distance from the origin, i.e. center of the aperture; A is the aperture area; and is the normal vector to surface A . We note that Eq. (12) can be significantly simplified if the measurement is performed at the diffraction center, (0, 0, z) with z≫λ. For a simple instance, assume that we measure a far-field signal from a w×L-sized slit together with the normalizing aperture of same length L but different width wa. The far-field signal from the slit at (0, 0, z) can be written as
where and are the electric field at the aperture-air boundary (z=0) and its averaging over the aperture area A, respectively. In the same manner, the far-field signal from the normalizing aperture can be expressed as
Equations (13) and (14) give rise to
This reveals that the ratio of far-field signals, directly measurable quantities, is related to that of the FEs, introducing the scale factor β given by the ratio of w and wa. Here, in the limit of wa>λ, one can find that FE by the reference aperture is close to 1, giving rise to where Einc is the amplitude of incident field. This enables a further approximation of Eq. (15), implying that FE can be determined exclusively by quantities measurable in far-field:
Despite its simple expression, we stress that Eq. (16) was successfully used in various THz experimental studies to estimate FE with reasonable accuracy in comparison with the theoretical predictions .
In the presence of the substrate, the FE can be also estimated by the same scheme, taking into account of the transmission of waves at the interfaces of substrate. At first, as shown in Figure 12, consider the far-field measurement at the substrate side with incident THz wave impinging upon the metal from the top side. If the thickness of the substrate is sufficiently thick, the transmission amplitude of the incident THz waves through the substrate-only sample, Esub, can be approximated by the direct transmission with ignoring the multiple reflection effect. That is, Then, the far-field transmission amplitude normalized by Esub can be written as
Here, we used the result of Eq. (16). Since t is an experimentally measurable quantity, Eq. (17) says that FE with substrate sample can be also quantitatively estimated from the far-field signal by using the Kirchhoff integral method:
For the reversed case, i.e. far-field measurement at the top side with incident THz wave impinging from the substrate side, the same scheme can be applied. What is different is that, in that case, the near-field signal radiates to the far-field freely without disturbance by the substrate. Therefore, Eq. (17) for the reversed measurement can be rewritten as
Interestingly, for the reversed measurement, it is better to just get the FE without using the substrate normalization. Then, Eq. (18) will be simply be where t′ is the transmission amplitude without going through the substrate normalization process.
2.3 Resonant versus nonresonant field enhancement
What we have discussed so far is basically nonresonant field enabled by the funneling process of THz waves through the gap, which is also related to the “capacitive enhancement”. In contrast to such nonresonant FE, it is also extensively discussed that two-dimensional apertures such as circular or rectangular holes can support resonant FE when certain resonance conditions are met. In terms of strong FE, a rectangular hole, also known as slot antenna, is practically more suitable than other types of holes. This is because of the nature of resonance condition that usually requires a geometric size of the hole to be in wavelength-scale. A slot antenna having an ultrahigh aspect ratio of width w and length l (i.e. w/l≪1) can utilize resonant light-slot coupling together with capacitive enhancement, maximizing the FE.
There have been extensive investigations to obtain quantitatively accurate FE by single slots in metal , , , , . The description of FE is shown to be also quite simple. However, compared to the single-slit case, its behavior is completely different. Analytic calculation based on the coupled-mode theory predicts that the resonant FE by a single slot of width w and length l (shown in Figure 13) is , 
where the resonance condition is simply given by
In Eq. (19), x varies from 0 to w. By taking average of the FE over the slot area, we have
It should be noted that retrieval of FE from the far-field measurement by using the discussed Kirchhoff integral method is related to Eq. (21), not Eq. (19). It is interesting to see that the resonant FE by single slot is given by the aspect ratio of width and length. Clearly, compared to the nonresonant Eq. (21) shows distinctive behaviors.
This simplified FE can be broadly applied to the range of micron gaps to nanogaps. Enhanced electromagnetic field through subwavelength gaps have been intensively explored both theoretically and experimentally over ultra-broadband wavelength regime covering microwave , , THz , , , , IR , , and visible regime , . Especially subwavelength photonics has been focused on THz frequencies (0.1–10 THz, 0.03–3 mm), having taken advantage of their relatively large aspect ratio between the wavelengths and structure scales, in turn providing a colossal FE effect. The studies, unavoidably, have strongly relied on the development of the state-of-the-art fabrication technologies for engineering small structures, which have been greatly advanced in the past decade , , . As an earlier work on THz metallic gap structures, micrometer scale rectangular hole arrays were prepared using femtosecond laser machining technique, which guarantees wavelength length scale perforation onto metal thin film (typically several tens to hundreds μm) and somewhat less , , as shown in Figure 13A.
The resonance can be tuned by changing the length of the rectangular holes, l. Then, as expected, the confined THz field near the gap and FE value increase with decreasing of the hole width (Figure 13B, C and Figure 14). The FE value is independent of the thickness h. Although we discussed the resonant FE only with a single slot case mostly so far, we note that a single slit can also support two different types of resonances: Fabry-Perot-type resonance that appears when the thickness of the slit is comparable to the half-wavelength ,  and the so-called fractional resonance from an ultrathin single slit filled with negative permittivity media .
3 THz nonlinearity in few- and subnanometer gaps by quantum electron tunneling
Classical theory of electromagnetism, describing microscopic light-matter interactions in terms of effective dielectric constant, predicts the THz FE by nanogaps in metals, in the form of electric field distribution obtained as a solution of the required boundary conditions at interfaces of metal and free-space , , , , . The predicted result is quantitatively accurate especially when the gap size is a few nanometers or larger , , , and the enhancement by an infinitely-long nanogap is shown to increase with decreasing gap size up to the enhancement factor, It is also confirmed experimentally that this behavior is still valid even when the gap approaches the regime of the sub-skin-depth .
One immediate question that arises here is whether this increasing behavior continues to the subnanometer regime. Considering the zero-gap limit in which there is no FE and that gap-size-dependent FE changes continuously with varying gap, one can presume that there should be a critical turning point of the gap size at which the near-field strength eventually starts decreasing with even narrower gap. Classical theory predicts that, in the perfect electric conductor approximation widely used in the study of THz wave interaction with metals, there is no such decrease in the FE. Exceptionally, the decreased FE with finite conducting metal is mainly due to the evanescent decay of the fundamental slit waveguide mode that becomes more pronounced with narrower gap in lossy material , . Recently, experimental and theoretical studies demonstrated that when the gap size approaches a few nanometers or Angstrom regimens, FE can be strongly modified and the classical theory cannot provide a quantitative understanding without taking into account quantum effects , , , , , , . In this regime, numerical schemes such as conventional finite-element method (FEM) and FDTD based on far-field dielectric function of the material also should have an implementation of the quantum effect for accurate calculations of FE.
There are two major quantum effects that come into play when the gap size approaches the nanometer or Angstrom regime. Nonlocal plasmon response , , , , , , , , , , , , , ,  is an extensively discussed, quantum-mechanical counterpart of the classical local response theory, giving rise to a suppression of the local FE through a modification of effective boundaries of materials within the Fermi-wavelength scale (nm to Angstrom). We note that the nonlocal effect can be directly applied to the FEM and FDTD schemes by adapting quantum-corrected dielectric functions of the materials. It should be also noted that, in some extent, the nonlocal effect is related to the Landau plasmon damping , ,  that results in additional suppression of the plasmonic electric field in the proximity of metals .
In this section, we discuss the other important quantum effect, the electron tunneling , , , , through a potential barrier of nanogap at THz frequencies , , , incorporating recent studies. In the point of view of electrons, a gap in a metal plate can be assumed as a tunnel junction , , as shown in Figure 15A. Usually, the gap size for an efficient electron tunneling needs to be subnanometer for typical potential barriers of the gap, a few eV. As shown in Figure 15B, an intense electric field is required to pull down the barrier for tunneling. We will see that, thanks to the ability of the nanogap in metal to capture the intensive THz field, the THz quantum tunneling and consequent strong nonlinear behavior result both in subnanometer , ,  and even in super-nanometer gaps , .
3.1 Fabrication of wafer-scale nanogap arrays
A challenging issue for the experimental studies on THz plasmonics with few- or subnanometer gap is sample fabrication because of extremely high ratio of photon wavelength and gap size. To have a reliable THz field interaction with nanogap and consequent plasmonic behaviors such as giant local FE and nonlinearity, the length of the gap should be at least in the scale of photon wavelength. Therefore, in pursuit of sufficiently intense far-field signal, the sample should cover several mm2 area with nanogap arrays of ultrahigh aspect ratio. To realize such large-area sample with high-resolution patterning, various recipes have been introduced, e.g. fs-laser machining for microscale punctured structures , , , focused ion beam , , e-beam lithography , , and photolithography techniques , ,  for metamaterial structures and, finally, atomic layer deposition (ALD) for atomic-scale gap structures , , , . For nanoscale gap structures, photolithography can be considered as a promising fabrication method for mass production at the same time, as shown in Figure 16. The stepping method using patterned mask guarantees several hundreds of nanometers in gap width with hundreds of microns in gap length, required for resonant enhancement of THz field (Figures 13 and 14). Also, by varying the length of the gap, l, one can obtain frequency tunable THz resonators (Figure 17A), showing a clear relationship between l and the fundamental resonance frequency, fres (Figure 17B).
For the ultimately small subnanogaps, on the one hand, a spacer deposition method ,  provided strong reliability in terms of preventing collapsing of the gap and controllability of precise thickness, enabling a 1-nm gap uniformly formed with several 100 microns length.
Figure 18 illustrates the sample fabrication scheme introduced by Jeong et al. , an improved version of the previously introduced scheme by Chen et al.  (Figure 19). Firstly, on the 3 nm chromium (Cr)/100 nm Au/sapphire substrate, 30 nm Cr and 150 nm aluminum Al layers are patterned by using a standard photolithography and liftoff process. This double layer is a sacrificial layer that will be used for removing excess metals and making the structure planar. Then, because Al is resistant to ion beam, ion milling is applied to remove exposed Au area while Au underneath the Al/Cr layer is sustained. After that, the ALD of aluminum oxide (Al2O3) is applied to form a uniform clad on the whole structure with nanometer-scale thickness. After the deposition of a second Au layer with adhesive Cr layer, Al/Cr wet etching is applied to remove the overhanging Au and Al layers.
In the above recipe, the nanogap is defined by the ALD of Al2O3, providing controllability in nanometer accuracy. Based on spacer deposition and filling schemes, and by adding a step of sacrificial layer, this method provides high-throughput fabrication of nanogaps with an ultrahigh aspect ratio.
Shown in Figure 20 is fabrication of Angstrom gaps introduced by Bahk et al. . First, on a patterned 300-nm-thick Cu film on a quartz substrate, a single-layer graphene (SLG) is seamlessly grown to cover the sidewalls. Then, an additional thinner Cu layer is deposited on the same sample by thermal evaporation. Finally, the second Cu layer is selectively peeled off by using an adhesive tape: the final result is Cu-SLG-Cu composite as shown in Figure 20C.
3.2 Nonlinear THz response by light-induced electron tunneling through nano- and Angstrom-sized gaps
A basic strategy to realize the THz nonlinearity is utilizing deep subwavelength gap to induce huge electric field that consequently pulls down the barrier in the tunnel junction by the potential energy V=eEgapw, where e is the charge of electron, Egap is the induced electric field in the gap, and w is the gap size. Since the electric field normal to the metal surface is related to the surface charge, Egap can be quantitatively determined by the accumulated surface charges at the two metal walls in the gap. This implies that, once electron tunneling occurs, a further charge accumulation by the incident THz field is limited and the FE will be reduced. Also, because the pulling down of the potential barrier increases with stronger Egap, naturally more intense THz field leads to more electron tunneling through the potential barrier, giving rise to more pronounced nonlinear responses in local electric FE and transmission of THz waves through the gap.
As discussed above, THz nonlinear behaviors through the nanogaps have been introduced by recent experimental studies , , . In Ref. , arrays of nanogap are prepared based on the previously discussed ALD method with the sacrificial layer , . An electron microscope image of the nanogap is shown in Figure 21A. Four different samples with selected gap sizes (1.5 nm, 2 nm, 5 nm, and 10 nm) are prepared to examine the THz nonlinearity depending on the tunneling probability. Shown in Figure 21B is the cross-section images of a 1.5-nm nanogap taken by a scanning electron microscope (top) and a scanning transmission electron micrograph (bottom). Note that the Al2O3 layer, presented in Figure 21B as a white area, defines the gap size, and its thickness, i.e. gap size, is controllable by the number of ALD cycles. The main result of this study is shown in Figure 21C. The THz transmission dependent on both Egap and gap width clearly shows that the tunneling process works to reduce the THz transmission, corresponding to the reduced local electric field. Also, we note that reduced transmission in a nanogap with smaller width is more sensitive to an increase in Egap, which is an intuitive result from that the smaller gap allows electron tunneling with higher probability. Nevertheless, it is clear that for the super-nano-gaps, what is most important is not the transient voltage applied to the gap so much as the electric field on the gap.
A nonlinear THz transmission can be quantitatively studied based on the dielectric response of the gap associated with the modal expansion method. This is an effective medium model in which the contribution of tunneling is interpreted as the transient modification of an effective dielectric function in the gap. As shown in Figure 21C, tunneling yields the tunneling current density, J, redistributing the induced surface charges near the gap. For a time-harmonic THz wave with angular frequency, ω, the polarization P=i(J/ω)exp(−iωt) can be obtained, and consequently, the effective dielectric function ε can be described as
The explicit form of the current density, J, dependent on both the gap size and the dielectric function of filling medium, i.e. Al2O3 in this case, can be evaluated by the image force model . By applying Eq. (22) to the nanogap with a filling medium possessing effective dielectric function, the FE can be obtained in analytic ways that we discussed in Section 2. As shown in Figure 21C, the calculated results are in very good agreement with experimental results.
Recently, squeezing THz waves into much narrower gaps, Angstrom-size gaps, and consequent tunneling effect resulting in significant reduction of light transmission were demonstrated with the help of two-dimensional (2D) van der Waals (vdW) materials . A heterostructure of 2D SLG and an effective vertically aligned 3 Å gap in copper layer, formed by Cu/1.5 Å gap/C/1.5 Å gap/Cu double vdW gaps of a 5 mm length, has revealed extreme THz nonlinearity with unprecedented 97% reduction of the transmittance through the gap (Figure 22). This self-limiting, gigantic optical nonlinearity is achieved by a massive THz funneling through Angstrom gaps, strongly pulling down the gap barrier and boosting the electron tunneling.
We note that the discussed THz quantum tunneling spans the boundaries of the study on quantum effects in microscopic light-matter interaction: down to the low frequency of 0.1 THz, up to strong field amplitudes of 5 V/nm, with barrier width from Angstrom to super-nanometer regimes.
4 Nanostructure-based absorber integrated terahertz sensor applications
Since many intramolecular and intermolecular vibration modes of molecules, including nucleobase , , , amino acid , , , , and protein , , , exist at THz frequency range, THz technology has been considered a promising mean for detection and spectroscopy of such small biomaterials. Even bigger and more complex biological structures such as cell , , ,  and tissues ,  can be explored with THz spectroscopic tools as well. Unlike ultraviolet light or X-rays, the low photon energy of THz (1 THz=4 meV) is an advantage to use because of noninvasiveness and nonionizing manner , . Nevertheless, an extremely small absorption cross-section has limited widespread use in the THz regime. In most spectroscopic cases, chemical compound samples were prepared in closely packed pellet form to get reasonably big signal. Furthermore, bio or chemical samples in solution state have huge water absorption at the THz frequency, and thermal fluctuation at room temperature can hinder high-throughput applications. Sometimes, the experiments have been performed at a low temperature to rule out thermal fluctuation issues . In this respect, recently advanced THz detection technologies assisted by micro-to-nano patterned (gap) structures can provide an excellent solution to these fundamental , , , , , , , , , , , , . Giant absorption cross-section enhancement of molecules inside nanogap structures will be discussed in this section showing a completely new type of label-free detection methods in room temperature. Examples of recent works will show that various type of samples are detectable even in very low concentration, including chemicals, protein, microorganism, and even viruses.
4.1 Increased molecular absorption cross-section by THz nanogaps
Metal nanogap structures can enormously enhance and strongly localize the THz electric field at a hot spot, increasing the THz absorption coefficients of molecules . Using a single nanogap structure in gold, it has been demonstrated that the molecular cross-section and absorption coefficient are enhanced by a factor of Egap/Z0Hgap~103, where Hgap is the magnetic field inside the gap. The origin of increased absorption coefficient is a strong asymmetry between THz electric and magnetic FEs: over a thousand times for the electric field, but in order of one for the magnetic field inside the gap. By considering the cross-section where dS is the Poynting vector absorbed between the points z and z+dz along the path of a THz beam, N is the number of absorptive molecules per unit volume, and σ is the molecular absorption cross-section, molecular absorption can be calculated as follows: the Fermi Golden rule says that one molecule absorbs (Joule per second), where μ is the electric dipole moment of the molecule, ρ(ħω0) is the density of states, and ω0 and E are the resonant angular frequency and the local electric field of light, respectively. A volume, Adz (A is a surface area and dz is a differential thickness), can be defined, inside which the molecular absorption is occurring. By energy conservation, we have
The third line of Eq. (23) implies that the molecular absorption cross-section, σ, is sensitive to the electromagnetic environment of E2/S, which greatly increases in the nanogap relative to the vacuum, owing to the asymmetric electromagnetic enhancements. A single nanogap with varying gap width from 50 nm to 5000 nm was used as a launching pad for strongly enhanced and localized THz field and a sensing hot spot at the same time. The first tested sample was an RDX (1,3,5-trinitroperhydro-1,3,5-trizine) powder that has an absorption feature around 0.8 THz, attributed to a molecular conformation or a weak hydrogen bond between RDX molecules. Therefore, the sensing chip with a fundamental resonance at 0.87 THz was designed to selectively enhance the absorption frequency and used to detect RDX in a very low amount as shown in Figure 23.
The strongly increased molecular cross-section by >103 was translated into a colossal absorption coefficient of ~170,000 cm−1. Thereby, extremely small quantities about 40 ng (even 22 fg inside the gap) could be detected . A theoretical model was also introduced to explain this and is in good agreement with a reference . At this point, it should be noted that there are several critical parameters in designing the THz nanoslot antenna in terms of the target samples. First, the fundamental resonance frequency shift for the nanogap structures should be accounted for molecule detection. When bio and chemical compounds are filled inside the gap, their effective refractive index can significantly affect the measured THz spectrum. The resonance frequency can be determined by the following relation : where l is the length of the gap and n is the real part of the complex refractive index. This means that, because the refractive index is dependent on the covered (drop-casted or filled) dielectric sample material, so is the shifting behavior of the resonance frequency , , . Also, the gap width is a very important parameter related to the amount of the THz FE . As discussed in the second section, a narrower gap yields a stronger THz FE, resulting in the increase in detection sensitivity. The gap-size-dependent detection sensitivity can be directly observed in the measurement of an extremely small amount of pesticides . According to the measurements in Ref. , the observable minimum concentration level of targeted methomyl solution is determined by the gap size, and surprisingly, using 500 nm and 100 nm width gap antennas, 8 ppm and 10 ppb of methomyl at the measureable limit were detected, respectively. This observable limit value is essential for the label-free sensing , especially for extremely low concentration of chemical and biological residual substances.
4.2 THz nanogap sensor applications for chemical compound identification
As previously discussed, enhanced THz field by metallic nano structures can provide highly advanced sensing performance basically in label-free , noncontact, and noninvasive manners. THz optical characteristics of various biomaterials can be represented in terms of the dielectric response in the THz range. For example, the refractive indices for various saccharides can be extracted from the transmitted THz spectra in the range of 0.5–2.5 THz , . Using the fast Fourier transform, the THz signal in time domain is converted to the amplitude and phase spectra in the frequency domain. Then the complex optical constants can be calculated with the waveforms for reference (typically void aperture) and target sample as a film (or pellet) form as follows:
where Es(ω) is the amplitude of the transmitted signal through the sample and Er(ω) is the amplitude of the input signal through empty space occupied with the film. n(ω) and α(ω) are the real parts of the refractive index and absorption coefficient, respectively, and d is the thickness of the film. The power absorption extracted from the difference between the spectral amplitudes passing through the sample and reference is attributed to the imaginary part of the refractive index, κ(ω). The real part of the refractive index, n(ω), is obtained from the phase difference between two signals as
and the absorption coefficient,
where φr is the phase of the reference waveform, φs is the phase of the signal that passed through the sample, and λ is the wavelength. So, in conclusion, the extracted optical parameters from the time-domain THz signals can be used in identifying the substances. On the other hand, the optical characteristics including absorption and refractive index should be further modified in the nanogap antenna system.
To closely investigate how the enhanced electric field can contribute to the detection situation, FDTD numerical simulation was implemented. We adapted a nonuniform meshing tool with a 10-nm gird at minimum to describe a deep subwavelength thickness of the film. The detectable target samples were assumed as homogeneously dielectric and somewhat absorptive clads with a certain thickness (much less than the wavelength), having the complex refractive index as A · n+iB · κ, where A and B are constant values over the broadband frequency regime in the range of 1.0–3.0. Using the auxiliary differential equation method, the absorptive media were implanted in the gold nanogap antenna.
Because the directly transmitted THz field passes the absorptive medium with a trend of exponential decay, the reduction of the transmission can be described as where Tr (ω)=(Er(ω))2 and Ts(ω)=(Es(ω))2 are the transmittances through the nanoantenna without and with the clads, C is the transmittance ratio at the air to clad interface, k=2π/λ is the incidence momentum, and h is the thickness of the clad, proportional to the molecular concentration (Figure 24A). As the clad increases from 0.5 to 5.0 μm, the transmittance decreases at the resonance frequency, an indication of increased absorption (Figure 24B).
On the one hand, the transmittance can be affected by the complex refractive indices as plotted in Figure 25A. While absorption is accompanied by the change in the imaginary part of the refractive index, κ, the change in the real part of the refractive index, the resonance frequency is mainly affected by n. The various changes in transmission spectra, therefore, can be clear evidence to discriminate the species of the target sensing materials. As plotted in Figure 25B, the peak of transmittance clearly decreases in terms of the imaginary part of the refractive index, κ; meanwhile, the resonance frequency mainly related to the real part of the refractive index, n, shows a small change. However, it should be noted that a stronger absorption can give an appreciable change in the resonance frequency.
In Ref. , different types of saccharide molecules possessing different molecular vibration modes (e.g. glucose at 1.4 THz and sucrose and fructose at 1.7 THz) were clearly distinguished using different length of gap antenna based sensing chips with different fundamental resonance frequency as shown in Figure 26. Using glucose-antenna (fres=1.4 THz), only the glucose sample shows the sensitively changed transmission spectrum; meanwhile, other saccharide samples do not. In the imaging data performed with the fructose-antenna (fres=1.7 THz), a strongly changed image contrast was observed at the fructose sample dropped area.
Applying this concept to various biomolecules and chemical molecules with specific vibration modes at broadband THz frequency regime, target samples in unprecedentedly low concentration level can be detected. This is valid even in case of more complex biological systems including protein, cell, and tissues with no specific spectral features due to their superposition and broadening of the spectrum. Such relatively large biomaterials can be considered a combination of many proteins, with vibration modes with inhomogeneous broadening such that no recognizable absorption features exist in the THz spectrum . In those cases, especially, the spectral shift can be treated as a key parameter. As shown in Figure 27, the resonance frequency shift of the metamaterial by the fungicide treatments is interpreted in terms of the number density and the dielectric constants of the microorganisms inside a gap area.
Besides single resonance gap structures, ultra-broadband resonance filters composed of several gap antennas with log-periodically varied lengths and periods in their arrangement at the same time can be an excellent alternative, utilizing their field-enhancement advantage but freely applicable to any target frequency , . Such ultra-broadband sensing chip has a decided merit for unknown targets, in other word, samples without any obvious spectral features in THz fingerprinting. This is very useful to identify and quantify such virus samples as shown in Figure 28 . THz optical characteristics based on the transmission reduction and resonance frequency shift by the covered virus samples can be mapped for different subtypes of the viruses and their quantifications. The suggested sensitive and selective THz detection in the reference provides abundant optical information of measured viruses, suggesting quick and accurate monitoring and rapid diagnosis of viruses.
5 Discussions and outlook
Maxwell’s theory provided descriptions of light interaction with metal in terms of microscopic coupling between incident electromagnetic waves and charges residing in metal. This immediately allowed manipulation of the moving of charges by the external light, which is now developed to colossal field enhancement by accumulating charges near a narrow gap in metal. This huge field enhancement is in particular important in the THz spectral regime as it can partly compensate the relatively low average power of THz sources and small cross-sections . Nonlinear THz response , , , , ,  is a prime example utilizing intense THz field that boosts light-matter interaction , . Furthermore, it is very clear that integration of the huge field enhancement in a gap with other novel optical/electrical/plasmonic properties of materials , ,  or devices , , , , , , , , ,  can expand the practical use of the THz field. Graphene-integrated plasmonic system , , , , , ,  is one immediate example, enabled by atomically thin and electrically tunable nature of graphene (Figure 29). We also expect that, considering the reciprocity of electromagnetic radiation , the light-capturing ability of nanogap could be a key principle to extract and radiate THz waves from the integrated devices.
Although we have mainly discussed resonant and nonresonant field enhancements in fundamental types of gaps, i.e. slit and rectangle slot, there are many other opportunities to realize even higher field enhancement. For instance, coupling of two or more symmetric/asymmetric gaps can be considered to obtain field enhancement by symmetric and antisymmetric resonance modes , , . Also, localizations of incident/transmitted light before/after the gap can be incorporated to increase the field enhancement in the gap , , ,  (Figure 30). The other example is to make use of hybridization of a gap with plasmonic metastructures or metaelements , , , , , , , , , , , ,  that results in intensive localization of electric field in the form of surface waves such as vortex-plasmon and spoof surface plasmon modes.
In this review, we discussed both the fundamentals and applications of THz wave interaction with deep subwavelength nanostructures. The enormous THz field enhancement can be explained as a result of surface charge accumulation near the nanogaps, with narrower gaps supporting stronger electric field enhancement. The fundamentals of giant THz field enhancement at an infinitely long nanogap were introduced in Section 2. Specifically, nonresonant THz field enhancements by nanogaps with both sub- and super-skin-depth thicknesses were discussed with a simple model based on the boundary conditions of Maxwell’s equations. From this model, we obtained a surprisingly simple description of the field enhancement valid over many orders of magnitudes of conductivity of metal as well as many orders of magnitudes of the film thickness. In terms of the real-world measurement, we revisited the Kirchhoff integral formalism that enables quantitative estimation of near-field enhancement by typical far-field measurements. At the end of Section 2, resonant THz field enhancement by critically defined boundaries such as rectangular slot shapes was also discussed. In Section 3, more extreme cases with deep subwavelength structures from nanoscale to Angstrom scale were discussed with some example of plasmonic quantum effect and electron tunneling with consequent nonlinear behaviors. Enhanced THz nonlinear phenomena through nano- to Angstrom-sized gaps show a new pathway for the observation of unprecedented resonant and nonresonant changes in various. In Section 4, finally, a new type of THz molecule sensor based on the field enhancement via subwavelength structures, a promising sensing tool for chemistry, biology, and medical applications was introduced. As one of representative applications, ultrasensitive and highly selective THz molecule sensor has been suggested. The ultrasensitive THz molecule sensing mechanism follows the huge THz electric field enhancement via nanogap structures, leaving the magnetic field magnitudes almost intact but with very small spatial curvatures introduced, resulting in greatly increased absorption cross-section of the covered substances. Since many intramolecular and intermolecular vibration modes of molecules exist in the THz range, frequency controllable THz metamaterials can be used as special molecule targeted sensors. As a further step, it can be applied for real-time capturing of some of the physical dynamics prominent in biological systems.
We thank Dasom Kim for the COMSOL simulations of distributions of current density near the slit. We also thank Hyosim Yang for thin film direct transmittance discussions. This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP: NRF-2015R1A3A2031768). Minah Seo was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIP) (2016R1A2B2010858) and KIST intramural grant nos. 2E27270 and 2V05550. Ji-Hun Kang was supported in part by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2013R1A1A2011757).
Hoffmann MC, Monozon BS, Livshits D, Rafailov EU, Turchinovich D. Terahertz electro-absorption effect enabling femtosecond all-optical switching in semiconductor quantum dots. Appl Phys Lett 2010;97:231108. CrossrefGoogle Scholar
Upadhyaya P, Pramanik S, Bandyopadhyay S. Optical transitions in a quantum wire with spin-orbit interaction and its applications in terahertz electronics: Beyond zeroth-order theory. Phys Rev B 2008;77:155439. CrossrefGoogle Scholar
Williams CR, Andrews SR, Maier SA, Fernandez-Dominguez AI, Martin Moreno L, Garcia-Vidal FJ. Highly confined guiding of terahertz surface plasmon polaritons on structured metal surfaces. Nat Photon 2008;2:175–9. CrossrefGoogle Scholar
Fitzgerald AJ, Berry E, Zinovev NN, Walker GC, Smith MA, Chamberlain JM. An introduction to medical imaging with coherent terahertz frequency radiation. Phys Med Biol 2002;47:R67. CrossrefPubMedGoogle Scholar
Zimdars D, Valdmanis JA, White JS, et al. Technology and applications of terahertz imaging non-destructive examination: inspection of space shuttle sprayed on foam insulation. AIP Conf Proc 2005;760:570–7. CrossrefGoogle Scholar
Semashkin EN, Artyushkina TV. Operating range and all-weather capability of terahertz (0.1 THz) and gigahertz (3–33.3 GHz) radars on horizontal and oblique tracks. J Opt Technol 2015;82:430–5. CrossrefGoogle Scholar
Zimdars D, White JS, Stuk G, Chernovsky A, Fichter G, Williamson S. Security and non destructive evaluation application of high speed time domain terahertz imaging. In: 2006 Conference on Lasers and Electro-Optics and 2006 Quantum Electronics and Laser Science Conference, IEEE, May 21–26, 2006, pp. 1–2. Google Scholar
Xu KK, Zhang ZY, Yu Q, Wen ZY. J Disp Technol 2016;12: 115–21. Google Scholar
Fukunaga K, Hosako I. Innovative non-invasive analysis techniques for cultural heritage using terahertz technologyTechniques innovantes d’analyse non invasive du patrimoine culturel basées sur les technologies térahertz. C R Phys 2010;11:519–26. CrossrefGoogle Scholar
Kaliteevski MA, Brand S, Garvie-Cook J, Abram RA, Chamberlain JM. Terahertz filter based on refractive properties of metallic photonic crystal. Opt Express 2008;16:7330–5. CrossrefPubMedGoogle Scholar
Gao WL, Shu J, Reichel K, et al. High-contrast terahertz wave modulation by gated graphene enhanced by extraordinary transmission through ring apertures. Nano Lett 2014;14:1242–8. PubMedCrossrefGoogle Scholar
Astley V, Reichel K, Mendis R, Mittleman DM. Terahertz microfluidic sensing using a parallel-plate waveguide sensor. J Vis Exp 2012;30:e4304. Google Scholar
Lee JW, Seo MA, Kang DH, Khim KS, Jeoung SC, Kim DS. Terahertz electromagnetic wave transmission through random arrays of single rectangular holes and slits in thin metallic sheets. Phys Rev Lett 2007;99:137401. CrossrefPubMedGoogle Scholar
Stefanovich G, Pergament A, Stefanovich D. Electrical switching and Mott transition in VO2. J Phys: Condensed Matter 2000;12:8837. Google Scholar
Kang JH, Park QH, Lee JW, Seo MA, Kim DS. Perfect transmission of THz waves in structured metals. J Korean Phys Soc 2006;49:881–4. Google Scholar
Koo S, Kumar MS, Shin J, Kim D, Park N. Extraordinary magnetic field enhancement with metallic nanowire: role of surface impedance in Babinet’s principle for sub-skin-depth regime. Phys Rev Lett 2009;103:263901. CrossrefPubMedGoogle Scholar
Berry CW, Wang N, Hashemi MR, Unlu M, Jarrahi M. Significant performance enhancement in photoconductive terahertz optoelectronics by incorporating plasmonic contact electrodes. Nat Comm 2013;4:1622. CrossrefGoogle Scholar
Jeong Y-G, Paul MJ, Kim S-H, Yee K-J, Kim D-S, Lee Y-S. Large enhancement of nonlinear terahertz absorption in intrinsic GaAs by plasmonic nano antennas. Appl Phys Lett 2013;103:171109. CrossrefGoogle Scholar
Lange C, Maag T, Hohenleutner M, et al. Extremely nonperturbative nonlinearities in GaAs driven by atomically strong terahertz fields in gold metamaterials. Phys Rev Lett 2014;113:227401. CrossrefPubMedGoogle Scholar
Novitsky A, Ivinskaya AM, Zalkovskij M, Malureanu R, Jepsen PU, Lavrinenko AV. Non-resonant terahertz field enhancement in periodically arranged nanoslits. J Appl Phys 2012;112:074318. CrossrefGoogle Scholar
Toma A, Tuccio S, Prato M, et al. Squeezing terahertz light into nanovolumes: nanoantenna enhanced terahertz spectroscopy (NETS) of semiconductor quantum dots. Nano Lett 2015;15:386–91. CrossrefPubMedGoogle Scholar
Jeong J, Rhie J, Jeon W, Hwang CS, Kim DS. High-throughput fabrication of infinitely long 10 nm slit arrays for terahertz applications. J Infrared Millim Terahertz Waves 2015;36:262–8. CrossrefGoogle Scholar
Jeong YG, Paul MJ, Kim SH, Yee KJ, Kim DS, Lee YS. Large enhancement of nonlinear terahertz absorption in intrinsic GaAs by plasmonic nano antennas. Appl Phys Lett 2013;103:171109. CrossrefGoogle Scholar
Shu J, Qiu CY, Astley V, Nickel D, Mittleman DM, Xu QF. High-contrast terahertz modulator based on extraordinary transmission through a ring aperture. Opt Express 2011;19:26666–71. PubMedCrossrefGoogle Scholar
Werley CA, Fan KB, Strikwerda AC, et al. Time-resolved imaging of near-fields in THz antennas and direct quantitative measurement of field enhancements. Opt Express 2012;20:8551–67. CrossrefPubMedGoogle Scholar
Iwaszczuk K, Andryieuski A, Lavrinenko A, Zhang XC, Jepsen PU. Terahertz field enhancement to the MV/cm regime in a tapered parallel plate waveguide. Opt Express 2012;20: 8344–55. CrossrefGoogle Scholar
Fan F, Xu ST, Wang XH, Chang SJ. Terahertz polarization converter and one-way transmission based on double-layer magneto-plasmonics of magnetized InSb. Opt Express 2016;24:26431–43. PubMedCrossrefGoogle Scholar
Fan F, Gu WH, Chen S, Wang XH, Chang SJ. State conversion based on terahertz plasmonics with vanadium dioxide coating controlled by optical pumping. Opt Lett 2013;38:1582–4. CrossrefPubMedGoogle Scholar
Martin-Cano D, Quevedo-Teruel O, Moreno E, Martin-Moreno L, Garcia-Vidal FJ. Waveguided spoof surface plasmons with deep-subwavelength lateral confinement. Opt Lett 2011;36:4635–7. PubMedCrossrefGoogle Scholar
Takano K, Shibuya K, Akiyama K, Nagashima T, Miyamaru F, Hangyo M. A metal-to-insulator transition in cut-wire-grid metamaterials in the terahertz region. J Appl Phys 2010;107:024907. CrossrefGoogle Scholar
Bouwkamp CJ. On the diffraction of electromagnetic waves by small circular disks and holes. Philips Res Rep 1950;5: 401–22. Google Scholar
Naweed A, Baumann F, Bailey WA, Karakashian AS, Goodhue WD. Evidence for radiative damping in surface-plasmon-mediated light transmission through perforated conducting films. J Opt Soc Am B 2003;20:2534–8. CrossrefGoogle Scholar
Bravo-Abad J, Fernandez-Dominguez AI, Garcia-Vidal FJ, Martin-Moreno L. Theory of extraordinary transmission of light through quasiperiodic arrays of subwavelength holes. Phys Rev Lett 2007;99:203905. PubMedCrossrefGoogle Scholar
Bell PM, Pendry JB, Moreno LM, Ward AJ. A program for calculating photonic band structures and transmission coefficients of complex structures. Comput Phys Commun 1995;85:306–22. CrossrefGoogle Scholar
Baida FI, Van Labeke D. Near-field distribution of optical transmission of periodic subwavelength holes in a metal film. Phys Rev B 2003;67:155314. Google Scholar
Lin J, Oh SH, Nguyen HM, Reitich F. Volume polarization holographic recording in thick photopolymer for optical memory. Opt Express 2014;22:14402–10. Google Scholar
Yasuda H, Hosako I. Terahertz waveguide design for GaSb/AlGaSb quantum cascade laser. Jpn J Appl Phys 2008;47:1632–4. Google Scholar
Ordal MAB, Bell RJ, Jr., Alexander RW, Jr., Long LL, Querry MR. Optical properties of fourteen metals in the infrared and far infrared: Al, Co, Cu, Au, Fe, Pb, Mo, Ni, Pd, Pt, Ag, Ti, V, and W. Appl Optics 1985;24:4493–9. CrossrefGoogle Scholar
Ordal MAL, Long L, Bell RJ, et al. Optical properties of the metals Al, Co, Cu, Au, Fe, Pb, Ni, Pd, Pt, Ag, Ti, and W in the infrared and far infrared. Appl Optics 1983;22:1099–120. CrossrefGoogle Scholar
Giannini V, Berrier A, Maier SA, Sánchez-Gil JA, Rivas JG. Scattering efficiency and near field enhancement of active semiconductor plasmonic antennas at terahertz frequencies. Opt Express 2010;18:2797–807. PubMedCrossrefGoogle Scholar
Seo MA, Adam AJL, Kang JH, et al. Fourier-transform terahertz near-field imaging of one-dimensional slit arrays: mapping of electric-field-, magnetic-field-, and Poynting vectors. Opt Express 2007;15:11781–9. PubMedCrossrefGoogle Scholar
Kyoung JS, Seo MA, Park HR, Ahn KJ, Kim DS. Far field detection of terahertz near field enhancement of sub-wavelength slits using Kirchhoff integral formalism. Opt Commun 2010;283:4907–10. CrossrefGoogle Scholar
Park HR, Koo SM, Suwal OK, et al. High-performance organic charge trap flash memory devices based on ink-jet printed 6,13-bis(triisopropylsilylethynyl) pentacene transistors. Appl Phys Lett 2010;96:213107. CrossrefGoogle Scholar
Gorelick S, Guzenko VA, Vila-Comamala J, David C. Direct e-beam writing of dense and high aspect ratio nanostructures in thick layers of PMMA for electroplating. Nanotechnology 2010;21:295303. PubMedCrossrefGoogle Scholar
Tseng AA, Kuan C, Chen CD, Ma KJ. Electron beam lithography in nanoscale fabrication: recent development. IEEE Transactions on Electronics Packaging Manufacturing 2003;26:141–9. CrossrefGoogle Scholar
Meijer J, Du K, Gillner A, et al. Laser machining by short and ultrashort pulses, state of the art and new opportunities in the age of the photons. CIRP Annals – Manufacturing Technology 2002;51:531–50. CrossrefGoogle Scholar
Morozov GV, Maev RG, Drake GW. Green’s function analysis of electromagnetic waves in two-layered periodic structures with fluctuations in thickness. Phys Rev E Stat Nonlin Soft Matter Phys 2001;63(5 pt 2):056601. PubMedCrossrefGoogle Scholar
Nicorovici NA, McPhedran RC, Petit R. Efficient calculation of the Green’s function for electromagnetic scattering by gratings. Phys Rev E Stat Phys Plasmas Fluids Relat Interdiscip Topics 1994;49:4563–77. PubMedGoogle Scholar
Lalanne P, Hugonin JP. Numerical performance of finite-difference modal methods for the electromagnetic analysis of one-dimensional lamellar gratings. J Opt Soc Am A Opt Image Sci Vis 2000;17:1033–42. PubMedCrossrefGoogle Scholar
Xu KK, Liu SY, Sun WF, et al. Design and fabrication of a monolithic optoelectronic integrated Si CMOS LED based on hot-carrier effect. Ieee J Sel Top Quant 2016;22:2000508. Google Scholar
Filter R, Bosel C, Toscano G, Lederer F, Rockstuhl C. Nonlocal effects: relevance for the spontaneous emission rates of quantum emitters coupled to plasmonic structures. Opt Lett 2014;39:6118–21. CrossrefPubMedGoogle Scholar
Varas A, García-González P, Feist J, García-Vidal FJ, Rubio A. Quantum plasmonics: from jellium models to ab initio calculations. Nanophotonics 2016;5:409–26. Google Scholar
Qian H, Xiao Y, Lepage D, Chen L, Liu Z. Quantum electrostatic model for optical properties of nanoscale gold films. Nanophotonics 2015;4:413–8. Google Scholar
Xie HY, Ng MY, Chang YC. Analytical solutions to light scattering by plasmonic nanoparticles with nearly spherical shape and nonlocal effect. J Opt Soc Am A Opt Image Sci Vis 2010;27:2411–22. PubMedCrossrefGoogle Scholar
Toscano G, Raza S, Jauho AP, Mortensen NA, Wubs M. Modified field enhancement and extinction by plasmonic nanowire dimers due to nonlocal response. Opt Express 2012;20:4176–88. CrossrefPubMedGoogle Scholar
Tserkezis C, Stefanou N, Wubs M, Mortensen NA. Molecular fluorescence enhancement in plasmonic environments: exploring the role of nonlocal effects. Nanoscale 2016;8:17532–41. PubMedCrossrefGoogle Scholar
Fernandez-Dominguez AI, Wiener A, Garcia-Vidal FJ, Maier SA, Pendry JB. Transformation-optics description of nonlocal effects in plasmonic nanostructures. Phys Rev Lett 2012;108:106802. CrossrefPubMedGoogle Scholar
Curtet M, Bonnaud G. Landau damping of an electron plasma wave in a plasma with modulated density. Phys Rev E Stat Phys Plasmas Fluids Relat Interdiscip Topics 1999;60:R5052–5. Google Scholar
Paredes-Juarez A, Iakushev DA, Flores-Desirena B, Makarov NM, Perez-Rodriguez F. Landau damping of electromagnetic transport via dielectric-metal superlattices. Opt Lett 2015;40:3588–91. PubMedCrossrefGoogle Scholar
Ward DW, Statz ER, Nelson KA. Fabrication of polaritonic structures in LiNbO3 and LiTaO3 using femtosecond laser machining. Appl Phys A 2007;86:49–54. Google Scholar
Yoshioka K, Katayama I, Minami Y, et al. Real-space coherent manipulation of electrons in a single tunnel junction by single-cycle terahertz electric fields. Nat Photonics 2016;10:762–5. CrossrefGoogle Scholar
King MD, Ouellette W, Korter TM. Noncovalent interactions in paired DNA nucleobases investigated by terahertz spectroscopy and solid-state density functional theory. J Phys Chem A 2011;115:9467–78. CrossrefPubMedGoogle Scholar
Nishizawa J-I, Sasaki T, Suto K, et al. THz transmittance measurements of nucleobases and related molecules in the 0.4- to 5.8-THz region using a GaP THz wave generator. Opt Commun 2005;246:229–39. CrossrefGoogle Scholar
Yamaguchi M, Miyamaru F, Yamamoto K, Tani M, Hangyo M. Terahertz absorption spectra of L, D, and DL-alanine and their application to determination of enantiometric composition. Appl Phys Lett 2005;86:053903. CrossrefGoogle Scholar
Kim SJ, Born B, Havenith M, Gruebele M. Real-time detection of protein-water dynamics upon protein folding by terahertz absorption spectroscopy. Angewandte Chemie International Edition 2008;47:6486–9. CrossrefGoogle Scholar
Walther M, Fischer B, Schall M, Helm H, Jepsen PU. Far-infrared vibrational spectra of all-trans, 9-cis and 13-cis retinal measured by THz time-domain spectroscopy. Chem Phys Lett 2000;332:389–95. CrossrefGoogle Scholar
Zhang C, Liang L, Ding L, et al. Highly sensitive and selective sugar detection by terahertz nano-antennas. Appl Phys Lett 2016;108:241105. Google Scholar
Luo L, Ge C, Tao Y, et al. High-eflciency refractive index sensor based on the metallic nanoslit arrays with gain-assisted materials. Nanophotonics 2016;5:139–46. Google Scholar
Zanchetta G, Lanfranco R, Giavazzi F, Bellini T, Buscaglia M. Emerging applications of label-free optical biosensors. Nanophotonics 2017;6:627–45. Google Scholar
Chiavaioli F, Baldini F, Tombelli S, Trono C, Giannetti A. Biosensing with optical fiber gratings. Nanophotonics 2017;6:663–79. Google Scholar
Kang BJ, Baek IH, Lee SH, et al. Highly nonlinear organic crystal OHQ-T for efficient ultra-broadband terahertz wave generation beyond 10 THz. Opt Express 2016;24: 11054–61. CrossrefPubMedGoogle Scholar
Shinokita K, Hirori H, Nagai M, Satoh N, Kadoya Y, Tanaka K. Dynamical Franz–Keldysh effect in GaAs/AlGaAs multiple quantum wells induced by single-cycle terahertz pulses. Appl Phys Lett 2010;97:211902. CrossrefGoogle Scholar
Hughes S, Citrin DS. Creation of highly anisotropic wave packets in quantum wells: dynamical Franz-Keldysh effect in the optical and terahertz regimes. Phys Rev B 1999;59:R5288–91. CrossrefGoogle Scholar
Kang JH, Wang S, Shi ZW, Zhao WY, Yablonovitch E, Wang F. Goos-Hänchen shift and even-odd peak oscillations in edge-reflections of surface polaritons in atomically thin crystals. Nano Lett 2017;17:1768–74. CrossrefPubMedGoogle Scholar
Tao H, Landy NI, Bingham CM, Zhang X, Averitt RD, Padilla WJ. A metamaterial absorber for the terahertz regime: design, fabrication and characterization. Opt Express 2008;16: 7181–8. CrossrefPubMedGoogle Scholar
Jiang JL, Zhang X, Zhang W, et al. Polarized low-coherence interferometer based on a matrix CCD and birefringence crystal with a two-dimensional angle. Opt Express 2017;25:16867–78. Google Scholar
Zhao HL, Ren GJ, Liu F, Xin HP, Bai YB, Yao JQ. Tunable terahertz source via liquid crystal grating coated with electron beam excited graphene: a theoretical analysis. Opt Commun 2017;390:137–9. CrossrefGoogle Scholar
Peres NMR, Bludov YV, Ferreira A, Vasilevskiy MI. Exact solution for square-wave grating covered with graphene: surface plasmon-polaritons in the terahertz range. J Phys-Condens Mat 2013;25:125303. CrossrefGoogle Scholar
Zhao H, Guo Q, Xia F, Wang H. Two-dimensional materials for nanophotonics application. Nanophotonics 2015;4:128–42. Google Scholar
Valdivia-Valero FJ, Nieto-Vesperinas M. Enhanced transmission through subwavelength apertures by excitation of particle localized plasmons and nanojets. Opt Express 2011;19:11545–57. CrossrefPubMedGoogle Scholar
Lin RR, Xu YB, Liu HY, Lan S, Gopal AV. Strong localization of terahertz wave and significant enhancement in electric field achieved in U-shaped resonators with a large aspect ratio. Appl Phys Lett 2013;103:123505. CrossrefGoogle Scholar
About the article
Published Online: 2018-03-09
Published in Print: 2018-05-24
Citation Information: Nanophotonics, Volume 7, Issue 5, Pages 763–793, ISSN (Online) 2192-8614, DOI: https://doi.org/10.1515/nanoph-2017-0093.
©2018 Yonghui Tian and Jianhong Yang et al., published by De Gruyter, Berlin/Boston. This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License. BY-NC-ND 4.0