Interactions between light and conducting nanostructures can result in a variety of novel and fascinating phenomena. These properties may have wide applications, but their underlying mechanisms have not been completely understood. From calculations of surface charge density waves on conducting gratings and by comparing them with classical surface plasmons, we revealed a general yet concrete picture about coupling of light to free electron oscillation on structured conducting surfaces that can lead to oscillating subwavelength charge patterns (i.e., spoof surface plasmons but without the dispersion property of classical surface plasmons). New wavelets emitted from these light sources then destructively interfere to form evanescent waves. This principle, usually combined with other mechanisms (e.g. resonance), is mainly a geometrical effect that can be universally involved in light scattering from all periodic and nonperiodic structures containing free electrons, including perfect conductors. The spoof surface plasmon picture may provide clear guidelines for developing metamaterial-based nano-optical devices.
Deep Dive into Fundamental mechanism underlying subwavelength optics of metamaterials: Charge oscillation-induced light emission and interference.
Interactions between light and conducting nanostructures can result in a variety of novel and fascinating phenomena. These properties may have wide applications, but their underlying mechanisms have not been completely understood. From calculations of surface charge density waves on conducting gratings and by comparing them with classical surface plasmons, we revealed a general yet concrete picture about coupling of light to free electron oscillation on structured conducting surfaces that can lead to oscillating subwavelength charge patterns (i.e., spoof surface plasmons but without the dispersion property of classical surface plasmons). New wavelets emitted from these light sources then destructively interfere to form evanescent waves. This principle, usually combined with other mechanisms (e.g. resonance), is mainly a geometrical effect that can be universally involved in light scattering from all periodic and nonperiodic structures containing free electrons, including perfect conduc
The various novel and unusual optical properties of conducting nanostructures, such as anomalous diffraction from metallic gratings, enhanced light transmission through subwavelength slits or holes, light polarizing through wire grid polarizers, surface-enhanced Raman scattering, negative refraction imaging, etc, have attracted tremendous attention in recent years. 1,2,3,4,5 To date the coupling of light with surface plasmons (SPs) have been widely adopted to explain these anomalous phenomena. However, the SP picture elaborated in numerous case studies in the literature actually corresponds to a very general concept about coupling of electromagnetic (EM) waves to free electron oscillation on conducting surfaces that can generate evanescent EM wave modes. This big picture is correct without doubt, but it is too general for one to obtain a clear and straightforward understanding of the essential underlying mechanism. Due to this uncertainty, the SP-like wave modes have been usually assumed to be similar to the classical SPs (CSPs) on planar metal surfaces, 6 but this assumption is obviously challenged by the fact that (nearly) perfectly conducting structures that do not support CSPs still have similar but stronger anomalous light scattering properties. 7 Conductors with positive permittivity do not support CSPs either, but they can also exhibit light transmission anomalies. 8,9,10 (Extraordinary transmission through gratings can even occur for acoustic waves, 11 which is completely irrelevant to SPs.) Because of these contradictions, the origin of anomalous light scattering from metallic nanostructures is still being argued (e.g. Refs. 12-14).
Using modern computing techniques one may numerically solve Maxwell’s equations for various complicated structures, but such computations have been largely focused on the EM fields. Surprisingly, the detailed mechanisms of free electron oscillation have been almost completely ignored in the literature although they are known to play the essential role in the SP picture. Very recently, we have briefly reported our computations of surface charge density waves (SCDWs) and the role they play in the process of enhanced light transmission through slit and hole arrays. 9 In this paper, we give a detailed and comprehensive illustration of the basic mechanism about light emission and interference from incidentwave-driven free electron oscillations, demonstrate that it is involved in light scattering from all periodic and nonperiodic conducting structures (including perfect conductors), and thus establish a simple, concrete and universal spoof SP picture. This picture may provide solid guidelines for designing nanooptical devices by directing people to concentrate on the geometrical parameters of conducting nanostructures so as to control the locations, strengths, and interference of the charge oscillation-induced light sources.
To illustrate the main picture, we start from the well-known principle of Thomson scattering of x-rays by electrons, 15 in which the incident x rays (EM waves with wavelengths ∼ 0.1 nm) force the electrons in atoms (not necessarily free electrons) to oscillate with the same frequency. According to the fact that accelerating charges radiate, the oscillating electrons then emit new wavelets, which form the scattered waves. In principle, this effect also exists in the long wavelength range (say λ > 0.1 µm), where electrons still oscillate with the incident wave (giving rise to oscillating polarization of the atoms). However, since now λ is much larger than the atoms (∼ 0.1 nm), the net charge density averaged on the wavelength scale is zero in the bulk. Net polarization-induced charges do exist on surfaces (or interfaces), but for non-conducting materials where electrons are bound to atoms and cannot move freely, the formation of net oscillating charges is very small even on rough surfaces.
A metal has free electrons, which move/oscillate easily on the surface in response to external EM waves and thus may emit new wavelets. But first note that a CSP corresponds to a surface-bound mode on the metal. If the oscillating charges emit light, how can the CSP be non-radiative? To clarify this ambiguity, let us see the Otto geometry in Fig. 1(a) as an example. 6 At a specific incident angle θ sp [greater than the critical angle arcsin(1/n p ) of the prism-vacuum interface], the incident wave can excite a CSP, which is a sinusoidal SCDW on the metal surface with a wavevector
where ε c is the permittivity of the metal and K = 2π/λ (λ the incident wavelength in vacuum). Here θ sp must satisfy n p K sin θ sp = Re(k sp ), where n p is the refractive index of the prism. Under this condition, the incident energy is largely transferred to the CSP, giving rise to a reflection dip, as can be proved by Fresnel theory. 6 Note that CSPs can be activated only on metals with Re(ε c ) < 0 [and meanwhile Im(ε c ) being small]. 6 The reason is that under this condition, the spati
…(Full text truncated)…
This content is AI-processed based on ArXiv data.
