Quantitative limits of electromagnetic energy tunneling through subwavelength ε-near-zero channels and bends: loss, breakdown, nonlocality, and three-dimensional implementation

Materials with relative permittivity near zero (ε-near-zero, ENZ) allow electromagnetic energy to tunnel through waveguide channels and bends whose transverse dimensions are far below the free-space wavelength, because the near-zero phase velocity forces the magnetic field inside an ENZ region to be spatially uniform, making the transmission nearly independent of the channel geometry. The originating 2006 theory (Silveirinha and Engheta) left a coherent set of practical and quantitative issues open: the lower bound on the minimal channel width set by material loss, dielectric breakdown, and conductor loss (the internal electric field grows roughly as $1/d$ as the channel thickness $d$ shrinks); the effects of finite wire density, interface discretization, and spatial dispersion (nonlocality) in wire-metamaterial implementations and their impact on tunneling frequency, bandwidth, and efficiency; the strongly oscillatory resonances and geometry-dependent singular field enhancements that appear below the plasma frequency where the real part of the permittivity is negative; and the extension to fully three-dimensional geometries and to optical frequencies where metals are not perfect conductors. The effect itself has been verified experimentally and extended theoretically, but the general quantitative closure of the loss/breakdown trade-off on minimal channel width and a dedicated rigorous nonlocal theory of wire-medium ENZ supercoupling remain open.

Background

In a homogeneous medium with time dependence $e^{-i\omega t}$, the wave equation for the magnetic field component $H_z$ (transverse-magnetic, TM, polarization in two dimensions) contains the relative permittivity $\varepsilon$ in the operator $\nabla\cdot(\varepsilon^{-1}\nabla H_z)+k_0^2\mu H_z=0$. When $\varepsilon\to 0$ (an ε-near-zero, ENZ, material — for example a Drude-type plasma at its plasma frequency, or a rectangular waveguide operated exactly at the cutoff of its dominant mode, which mimics $\varepsilon=0$ for the effective 2D problem), the magnetic field inside a simply connected ENZ region becomes spatially uniform. Silveirinha and Engheta [Phys. Rev. Lett. 97, 157403 (2006)] exploited this to solve, in closed form, the scattering of two parallel-plate perfect-electric-conductor (PEC) waveguides of heights $a_1$ and $a_2$ joined by an arbitrarily shaped, subwavelength ENZ channel of cross-sectional area $A_p$: the reflection coefficient at the input plane is exactly $\rho=((a_1-a_2)+ik_0\mu_{r,p}A_p)/((a_1+a_2)-ik_0\mu_{r,p}A_p)$, where $k_0=\omega\sqrt{\varepsilon_0\mu_0}$ and $\mu_{r,p}$ is the relative permeability of the channel. When the waveguide heights match and the dimensionless combination $k_0\mu_{r,p}A_p/(a_1+a_2)$ is small, the reflection is arbitrarily small, so energy 'tunnels' through an electrically tiny channel with nearly complete transmission, independent of its length and shape — a phenomenon the same authors later named supercoupling. The price is that energy conservation forces the electric field amplitude inside the channel to grow roughly as $1/d$, where $d$ is the channel thickness. The same paper extended the result to anisotropic ENZ channels (the component $\varepsilon_{xx}$ along the plates is the one that must approach zero) and proposed realizing the ENZ response at microwaves with an array of metallic wires — a wire medium, whose connected-wire geometry has a Drude-like effective plasma frequency. Full-wave simulations of the wire-medium bend reproduced the predicted tunneling dip but with small frequency shifts attributed to finite wire spacing, interface (finite-size) effects, and spatial dispersion — the nonlocal dependence of the effective permittivity on the wave vector that is intrinsic to structured wire media. The companion detailed theory [Phys. Rev. B 76, 245109 (2007)] further proved that losses in the metallic waveguide walls enter the transmission only at second order (so dielectric loss in the ENZ material dominates), analyzed the field singularities at metal corners joining materials of positive and negative permittivity, and proposed 'simulated plasma' (cutoff-waveguide) implementations. Subsequent work verified the effect experimentally at microwaves using a waveguide near cutoff (which emulates ENZ without any metamaterial), demonstrated reflectionless sharp bends, mode-matched analyses, and nonlinear control of the tunneling resonance, and generalized the ideal-fluid picture of near-zero-index propagation. What was never supplied, for the lossy, structured, three-dimensional, and optical settings, is the quantitative closure of the limitations the source paper itself flagged — the subject of this problem. Terminology used below: 'dielectric breakdown' is the failure of the filling material under a too-large electric field; 'spatial dispersion' (nonlocality) means the effective permittivity depends on the wave vector, producing additional propagating waves that require additional boundary conditions beyond the standard Maxwell continuity conditions at wire-medium interfaces; 'quasistatic resonance' means a resonance whose physics is governed by the electrostatic (Laplace-type) problem rather than by retardation.

Problem Statement

For the ε-near-zero (ENZ) tunneling and supercoupling effect — near-total, geometry-independent transmission of electromagnetic energy through subwavelength channels and bends filled with an ENZ material — establish the quantitative limits of the effect in realistic implementations, as flagged by the source theory: (1) determine the lower bound on the minimal achievable channel width set by material loss, dielectric breakdown, and conductor loss, given that the internal electric field amplitude grows roughly as $1/d$ in the channel thickness $d$, and quantify these trade-offs for realistic materials and maximum usable field intensities; (2) give a rigorous analysis of the spatial-dispersion (nonlocal) and finite-size/interface effects in structured implementations such as wire media — including the frequency shifts they cause — and of their impact on the tunneling bandwidth and efficiency; (3) characterize and control the strongly oscillatory resonant behavior and the geometry-dependent singular field enhancements near metal edges that arise below the plasma frequency where $\mathrm{Re}\{\varepsilon\}<0$, in the presence of losses; and (4) extend the two-dimensional results to fully three-dimensional geometries beyond the classes already treated, including the role of nonideal conductor behavior at optical frequencies where metals are not perfect conductors. These four sub-goals together constitute the research target. Quantify, for a given real material and field-intensity budget, the achievable transmission, bandwidth, and minimal channel size of an ENZ supercoupling interconnect, in closed form or by quantitatively validated, reproducible criteria with specified material models, valid over an explicitly stated range of material parameters and field intensities, and with any bound stated in terms of material parameters rather than as a formally true but non-constraining inequality.

References

1. Mário Silveirinha and Nader Engheta, 'Tunneling of Electromagnetic Energy through Subwavelength Channels and Bends using ε-Near-Zero Materials', Phys. Rev. Lett. 97, 157403 (2006). DOI: 10.1103/PhysRevLett.97.157403, https://doi.org/10.1103/PhysRevLett.97.157403
2. M. Silveirinha and N. Engheta, 'Theory of Supercoupling, Squeezing Wave Energy, and Field Confinement in Narrow Channels and Tight Bends Using Epsilon-Near-Zero Metamaterials', Phys. Rev. B 76, 245109 (2007). arXiv:0705.2612, DOI: 10.1103/PhysRevB.76.245109, https://arxiv.org/abs/0705.2612
3. Brian Edwards, Andrea Alù, Michael E. Young, Mário Silveirinha, Nader Engheta, 'Experimental Verification of Epsilon-Near-Zero Metamaterial Coupling and Energy Squeezing Using a Microwave Waveguide', Phys. Rev. Lett. 100, 033903 (2008). DOI: 10.1103/PhysRevLett.100.033903, https://doi.org/10.1103/PhysRevLett.100.033903
4. B. Edwards, A. Alù, M. G. Silveirinha, N. Engheta, 'Reflectionless sharp bends and corners in waveguides using epsilon-near-zero effects', J. Appl. Phys. 105, 044905 (2009). DOI: 10.1063/1.3074506, https://doi.org/10.1063/1.3074506
5. Silveirinha and Engheta, 'Squeezing electromagnetic energy through narrow 3-D coaxial channels of arbitrary shape filled with Epsilon-Near-Zero materials', Proc. 2007 IEEE Antennas and Propagation Society International Symposium, pp. 3316–3319. DOI: 10.1109/APS.2007.4396246, https://doi.org/10.1109/APS.2007.4396246
6. David A. Powell, Andrea Alù, Brian Edwards, Ashkan Vakil, Yuri S. Kivshar, Nader Engheta, 'Nonlinear control of tunneling through an epsilon-near-zero channel', Phys. Rev. B 79, 245135 (2009). DOI: 10.1103/PhysRevB.79.245135, https://doi.org/10.1103/PhysRevB.79.245135
7. Yongmei Pan and Shanjia Xu, 'Rigorous mode-matching analysis of energy tunneling through an ultranarrow epsilon-near-zero channel', Microw. Opt. Technol. Lett. 51, 163–166 (2009). DOI: 10.1002/mop.23994, https://doi.org/10.1002/mop.23994
8. Christin David, N. Asger Mortensen, Johan Christensen, 'Perfect imaging, epsilon-near zero phenomena and waveguiding in the scope of nonlocal effects', Sci. Rep. 3, 2526 (2013). DOI: 10.1038/srep02526, https://doi.org/10.1038/srep02526
9. Silveirinha, 'Additional boundary conditions for nonconnected wire media', New J. Phys. 11, 113016 (2009). DOI: 10.1088/1367-2630/11/11/113016, https://doi.org/10.1088/1367-2630/11/11/113016
10. Maslovski and Silveirinha, 'Nonlocal permittivity from a quasistatic model for a class of wire media', Phys. Rev. B 80, 245101 (2009). DOI: 10.1103/PhysRevB.80.245101, https://doi.org/10.1103/PhysRevB.80.245101
11. Yakovlev, Silveirinha, and Hanson, 'Effective Local Permittivity Model for Nonlocal Wire Media', IEEE Trans. Antennas Propag. 68, 2926–2936 (2020). DOI: 10.1109/TAP.2019.2951129, https://doi.org/10.1109/TAP.2019.2951129
12. Iñigo Liberal, Michaël Lobet, Yue Li, Nader Engheta, 'Near-zero-index media as electromagnetic ideal fluids', Proc. Natl. Acad. Sci. USA 117, 24050–24054 (2020). DOI: 10.1073/pnas.2008143117, https://doi.org/10.1073/pnas.2008143117
13. Yichao Liu, Fei Sun, Yibiao Yang, Zhihui Chen, Jianzhong Zhang, Sailing He, Yungui Ma, 'Broadband Electromagnetic Wave Tunneling with Transmuted Material Singularity', Phys. Rev. Lett. 125, 207401 (2020). DOI: 10.1103/PhysRevLett.125.207401, https://doi.org/10.1103/PhysRevLett.125.207401
14. Wendi Yan, Peihang Li, Jiarui Liu, Kaifeng Li, Pengyu Fu, Shuyu Wang, Mingzhe Hu, Yue Li, 'Ohm's law of electromagnetic ideal fluids: impedance-governed supercoupling in complex near-zero-index networks', Nat. Commun. 17, 7131 (2026). DOI: 10.1038/s41467-026-73657-1, https://doi.org/10.1038/s41467-026-73657-1