Metagratings on Low-Cost Substrates for Efficient Anomalous Reflection: Addressing Dielectric Loss

TL;DR

Metagratings (MGs) offer precise beam steering via Floquet harmonics, but practical, low-cost substrates like FR4 introduce dielectric losses that degrade efficiency. The authors develop a loss-aware semi-analytical framework based on a two-meta-atom-per-period MG and an equivalent-circuit model to capture dielectric and conductor losses, enabling synthesis of high-efficiency anomalous reflection. The method yields actionable design steps and a parameter-extraction workflow, and is validated by full-wave simulations and a FR4 MG prototype that achieves an anomalous-reflection efficiency above 0.8 at 20 GHz. This demonstrates the feasibility of low-profile, cost-effective MGs for beam manipulation and absorber applications across the EM spectrum.

Abstract

We present a theoretical framework and practical methodology for designing high-efficiency metagratings (MGs), sparse periodic arrangements of subwavelength polarizable particles (meta-atoms), on low-cost dielectric substrates with non-negligible losses. The formulation incorporates these losses and exploits multiple degrees of freedom to optimize beam manipulation efficiency within a simple realistic printed-circuit-board (PCB) configuration. Importantly, the various loss mechanisms are analyzed using a judiciously devised equivalent circuit model, providing insights on their respective contributions. We validate our theory by designing, fabricating, and experimentally characterizing an efficient FR4-based anomalous reflection PCB MG, demonstrating good agreement between analytical predictions, full-wave simulations, and laboratory measurements. This work opens avenues for realizing efficient, low-profile, beam manipulation devices at reduced cost, offering practical solutions to mitigate loss limitations in diverse material sets across the electromagnetic spectrum.

Figures (13)

• Figure 1: MG physical configuration. (a) Two different grids of loaded strips lie on a dielectric substrate backed by PEC and shifted by distance $d$ from each other. The MG excited by a TE polarized plane wave with an angle of incidence of $\theta_\mathrm{in}$. (b) Top view of the MG, consisting of two different load impedance per unit length $\tilde{Z}_{1}$ and $\tilde{Z}_{2}$ per period, featuring periodicities $\Lambda$ and $l$ along $y$ and $x$, respectively. (c) Zoom in on a single meta-atom, featuring a printed capacitor of width $W$ with copper traces of dimensions $s$ and $w$ and thickness $t$.
• Figure 2: (a) Physical configuration for the $\tilde{Z}\textcolor{black}{_T}(W)$ look-up table extraction (Section \ref{['subsec:dielectric_loss']}). A single meta-atom with printed capacitor load of width $W$ (swept in a full-wave solver) is placed in a ($l\times\Lambda'$)-periodic configuration, defined on a $h'$-thick PEC-backed FR4 substrate. The impedance extraction procedure involves recording the reflection coefficient $S_{11}$ at the plane $z=-2\lambda$ for a given $W$ using a full-wave solver, deducing the effective induced current $I$ via (\ref{['Eq:Current_S11']}), and finally evaluating the effective $\tilde{Z}_{T}(W)$ based on Ohm's law (similar to \ref{['Eq:I0_I1']}). (b) The extracted relation between the meta-atom geometry and the associated distributed load impedance $\tilde{Z}_{T}(W)=\tilde{R}_{T}(W)+j\tilde{X}_{T}(W)$at the operating frequency $f=20$ GHz.
• Figure 3: Equivalent circuit of the MG printed load (corresponding to the effective distributed load impedance $\tilde{Z}_{T}$) of Fig. \ref{['fig:Setup']}(b) and (c). The printed load geometry incorporates two distinct RLC circuits: a series configuration and a parallel configuration (Section \ref{['subsec:circuit_model']}). The series RLC circuit (formed by $C_s$, $L$, $R_s$, and $R_c$) exhibits greater prominence within the lower range of $W$ values ($W<W_{s}$), while the parallel RLC circuit (formed by $C_p$, $L$, $R_p$, and $R_c$) becomes more dominant as W increases.
• Figure 4: Analysis of various loss factors contributing to the effective printed load impedance $\tilde{Z}_{T}(W)=\tilde{R}_{T}(W)+j\tilde{X}_{T}(W)$ corresponding to Figs. \ref{['fig:ExtrConfig1andLUT']} and \ref{['fig:EqvCirc']}. (a) $\tilde{R}_{T}(W)$ as extracted from full-wave simulations following Section \ref{['subsec:dielectric_loss']} for the configuration in Fig. \ref{['fig:ExtrConfig1andLUT']} with different conductor properties, dielectric substrates, and loss level: lossless FR4 $\varepsilon_{2,\mathrm{r}}'=4.4$ as substrate with PEC (solid blue) or realistic finite-conductivity copper (solid black) as meta-atom conductors; FR4 substrates with loss tangent varying from $\tan\delta=0.01$ (dot-dashed green), through$\tan\delta=0.02$ (dotted green), to $\tan\delta=0.04$ (dashed green), considering realistic copper conductors; and a reference configuration with only conductor loss (solid red), featuring copper meta-atoms suspended in vacuum ($\varepsilon_{2,\mathrm{r}}'=1$). (b) Relative resistance between the extracted $\tilde{R}_{T}(W)$ from (a) for $\tan\delta=0.01$ (dash-dotted green) and $\tan\delta=0.02$ (dotted green) with respect to the case $\tan\delta=0.04$. For all cases, realistic copper is used and $\varepsilon_{2,r}'=4.4$. (c) Similarly extracted $\tilde{X}_{T}(W)$ for representative scenarios from (a) with realistic copper conductors suspended in vacuum (solid red), or defined on a lossless (solid black) or realistic (dashed green) FR4.
• Figure 5: (a) Effective load resistance (blue solid curve, left $y$ axis) and reactance (red solid curve, right $y$ axis) as a function of capacitor widtth $W$, extracted from full-wave simulations following the procedure in Section \ref{['subsec:dielectric_loss']} for the configuration in Fig. \ref{['fig:ExtrConfig1andLUT']} in an extended range of capacitor widths (cf. Section \ref{['par:parallel_RLC']}). (b) Effect of substrate loss tangent on the effective printed load resistance $\tilde{R}_{T}(W)$ extracted as in (a) for the same extended width. Results are shown for realistic copper meta-atom conductors and FR4 substrates $\varepsilon_{2,r}'=4.4$ with loss tangent varying from $\tan\delta=0.01$ (green), through realistic $\tan\delta=0.02$ (blue), to $\tan\delta=0.04$ (red).