[Paper Review] Two new type surface polaritons excited into nanoholes in metal films

V. N. Minasyan, В. Н. Самойлов|arXiv (Cornell University)|Dec 23, 2009

Plasmonic and Surface Plasmon Research6 references3 citations

TL;DR

This paper proposes a novel theoretical framework in which the metal-air interface is modeled as a dielectric metal skin composed of neutral dipoles, leading to the prediction of two new types of surface polaritons in nanoholes. By quantizing Maxwell’s equations in this medium, the authors derive light boson quasi-particles with spin one and an effective mass of $2.5 \times 10^{-5}m_e$, explaining enhanced optical transmission and absorption anomalies via resonant field localization in subwavelength holes, confirmed by agreement with experimental transmission peaks at $\lambda_0 = 1.24\mu\text{m}$.

ABSTRACT

First, we argue that the smooth metal-air interface should be regarded as a distinct dielectric medium, the skin of the metal. The existence of this metal skin leads to theoretical explanation of experimental data on the excitation of electromagnetic surface shape resonances in lamellar metallic gratings by light in the visible to near-infrared range. Surface polaritons have been observed in reflection modes on metallized gratings where the electric field is highly localized inside the grooves (around 300-1000 times larger than intensity of incoming optical light). Here we present quantized Maxwell's equations for electromagnetic field in an isotropic homogeneous medium, allowing us to solve the absorption anomaly property of these metal films. The results imply the existence of light boson particles with spin one and effective mass $m= 2.5\cdot 10^{-5} m_e$. We also show the presence of two new type surface polaritons into nanoholes in metal films.

Motivation & Objective

• To resolve the long-standing absorption anomaly in metal films perforated with subwavelength holes, where optical transmission exceeds expectations.
• To explain the experimentally observed 300–1000-fold enhancement of electric field intensity inside nanohole grooves.
• To propose a new theoretical model where the metal-air interface is treated as a distinct dielectric medium (metal skin) with dipolar character.
• To derive the existence of two new surface polaritons via quantized electromagnetic field theory in this dielectric medium.
• To reconcile experimental transmission spectra with theoretical predictions using Bessel function dispersion relations in cylindrical resonators.

Proposed method

• Model the metal surface as a dielectric skin of neutral dipoles, each consisting of a bound electron and ion connected by a spring with resonance frequency $\omega_0$.
• Apply quantized Maxwell’s equations to the dielectric medium, introducing a Bose-gas of massive photons with effective mass $m = 2.5 \times 10^{-5}m_e$ via a field quantization scheme.
• Derive modified wave equations for electric and magnetic fields $\vec{E}_0$ and $\vec{H}_0$ that satisfy Maxwell’s equations in the dielectric medium.
• Use Bessel functions $J_0$ and $J_1$ to model standing wave conditions in cylindrical nanoholes, with boundary conditions at the metal skin-vacuum interface.
• Formulate dispersion equations linking the hole geometry ($d = 0.75\mu\text{m}$) and dielectric constant $\varepsilon$ to resonance frequencies.
• Apply the limit $\varepsilon \to \infty$ at $\omega = \omega_0$ to identify resonant transmission conditions, leading to $J_1(\omega_0 d / c) = 0$.

Experimental results

Research questions

• RQ1 Can the absorption anomaly in metal films with subwavelength nanoholes be explained by a new model of the metal surface as a dielectric skin rather than a free-electron gas?
• RQ2 What are the electromagnetic properties of light in a dielectric medium composed of bound electron-ion dipoles at the metal surface?
• RQ3 Do quantized electromagnetic fields in such a medium give rise to new surface polariton modes with finite effective mass?
• RQ4 How does the geometry of nanoholes influence the resonant transmission of light in the visible to near-infrared range?
• RQ5 Can the theoretical prediction of resonance frequency $\omega_0 = 3.8c/d$ reproduce the observed experimental transmission peak at $\lambda_0 = 1.24\mu\text{m}$?

Key findings

• The metal skin model, composed of neutral dipoles with resonance frequency $\omega_0$, leads to a divergent dielectric constant $\varepsilon \to \infty$ at resonance, explaining extreme field localization.
• Two new types of surface polaritons are predicted, with energies $\chi_{\vec{k}}$ and $\eta_{\vec{k}}$, which are excited in nanoholes and exhibit field intensities 300–1000 times greater than the incident light.
• The resonance frequency $\omega_0 = 3.8c/d$ is derived from the condition $J_1(\omega_0 d / c) = 0$, yielding a theoretical resonance wavelength $\lambda_0 = 1.24\mu\text{m}$.
• This predicted wavelength matches experimental transmission peaks observed in metal films with $d = 0.75\mu\text{m}$, confirming the model's validity.
• The effective mass of the light boson quasi-particles is calculated as $m = 2.5 \times 10^{-5}m_e$, consistent with the quantization scheme applied to the electromagnetic field in the dielectric medium.
• Phase velocities $v_e = c\varepsilon^{3/2}$ and $v_h = c\varepsilon^{1/2}$ exceed $c$, indicating a breakdown of standard phase velocity interpretation and explaining the absorption anomaly.