Ridge Waveguide Geometry

• Ridge waveguide geometry is defined by a ridge fabricated on or within a substrate, enabling lateral and vertical field confinement through index contrast or metallic boundaries.
• The waveguide’s design parameters such as ridge shape, dimensions, and material composition critically influence modal dispersion, loss mechanisms, and phase-matching conditions.
• Optimized ridge designs support advanced applications including resonant cavities, nonlinear frequency conversion, and leaky-wave antennas through precise fabrication and mode engineering.

A ridge waveguide is a guided-wave optical or electromagnetic structure in which a higher-index or conducting ridge is fabricated on, in, or above a lower-index planar substrate. The ridge induces lateral and vertical confinement of electromagnetic modes by leveraging refractive index contrast or metallic boundary conditions, supporting single-mode or multimode propagation with tight field localization. Ridge waveguides underpin numerous applications in integrated photonics, quantum optics, nonlinear optics, millimeter-wave engineering, and surface plasmonics, with geometry-driven performance spanning mode confinement, loss, phase-matching, and spectral selectivity.

1. Geometric Foundations and Taxonomy

The defining geometry of a ridge waveguide is a protuberance (rectangular, elliptical, sinusoidal, or otherwise shaped) that modulates the local electromagnetic environment, typically sitting atop, or being etched into, a slab or substrate of contrasting permittivity or conductivity.

Representative cross-sectional parameterizations include:

• Rectangular Ridge: width $w$, height $h$, vertical sidewalls or specified taper/inclination, on a substrate of thickness $t_{\mathrm{sub}}$ and refractive index $n_{\mathrm{sub}}$ (Gür et al., 2022, Pan et al., 2011).
• Elliptical Ridge: semi-axes $a$ (half-width), $b$ (half-height), with top boundary $(x/a)^2 + (y/b)^2 = 1$ [$w=2a$, $h=b$]; apex curvature $\kappa = b/a^2$ (Pan et al., 2012).
• Sinusoidal/Modulated Ridge: amplitude $h$0, width $h$1, spatial period $h$2; used for leaky-wave or radiative applications (Mohammad-Ali-Nezhad et al., 2018).
• Multilayer (Metamaterial) Ridge: stack of alternating sub-wavelength layers, parameterized by fill factor $h$3, unit-cell thickness $h$4, and overall width/height $h$5 (Sifat et al., 2016).

Material systems span undoped semiconductors (Si, GaAs, LiNbO$h$6, AlGaAs, KTP), noble metals (Ag/Au), all-dielectric metamaterials (alternating Si-SiO$h$7), superconductors (NbTiN), and hybrid glass–metallic substrates (Pan et al., 2011, Pan et al., 2012, Sifat et al., 2016, Banys et al., 2021).

Ridge geometry defines the core guiding properties—modal index, effective cross-sectional area, propagation loss, and field overlap.

2. Modal Analysis and Dispersion Characteristics

2.1 Vectorial Eigenvalue Problem

The electromagnetic modes of a ridge waveguide are solutions to the vectorial eigenvalue problem,

$h$8

where spatially varying $h$9 defines the ridge, slab, and cladding/substrate regions. For plasmonic or hybrid structures, complex permittivity tensors and anisotropy are incorporated (EMT for multilayers) (Sifat et al., 2016).

2.2 Mode Coupling and Field Confinement

• Confinement is maximized by high index contrast, tighter ridge dimensions (reducing $t_{\mathrm{sub}}$0 and/or $t_{\mathrm{sub}}$1), and optimized apex curvature (elliptical/semi-circular tops for SPP devices).
• Dispersion: The effective mode index, $t_{\mathrm{sub}}$2, is a nonlinear function of $t_{\mathrm{sub}}$3, saturating toward bulk or wedge limits as the ridge widens. In elliptical nano-ridges, optimal $t_{\mathrm{sub}}$4 and propagation length $t_{\mathrm{sub}}$5 occur for semi-circular ridge tops ($t_{\mathrm{sub}}$6, $t_{\mathrm{sub}}$7) (Pan et al., 2012).
• Loss mechanisms: Ohmic loss dominates for metallic ridges (Ag, Au); radiation loss is critical for insufficiently confined dielectric modes—especially at smaller ridge heights or for large hole diameters in photonic Bragg mirrors (Gür et al., 2022, Pan et al., 2011).

Key analytical and semi-empirical relations:

• Flat-top plasmonic ridge: mode size $t_{\mathrm{sub}}$8 (with $t_{\mathrm{sub}}$9 the field decay constant), propagation length $n_{\mathrm{sub}}$0, figure-of-merit $n_{\mathrm{sub}}$1 (Pan et al., 2011).
• EMT for metamaterial ridge: transverse and longitudinal permittivities ($n_{\mathrm{sub}}$2, $n_{\mathrm{sub}}$3) determined via fill factor $n_{\mathrm{sub}}$4 for sub-wavelength stacking (Sifat et al., 2016).

3. Resonant and Cavity Ridge Configurations

3.1 Dielectric Ridge Resonators and BICs

High-$n_{\mathrm{sub}}$5 resonance and bound-state-in-continuum (BIC) phenomena emerge in dielectric ridge waveguides atop slabs. Quasi-TE and quasi-TM mode hybridization enables spectral sharpness with analytical loci defined by ridge width $n_{\mathrm{sub}}$6 and angle of incidence $n_{\mathrm{sub}}$7:

$n_{\mathrm{sub}}$8

with $n_{\mathrm{sub}}$9 derived from Fabry–Pérot resonance conditions (Bezus et al., 2018). Properly chosen $a$0 and angle selectivity yield ultra-high $a$1 for sensing, filtering, or nonlinear enhancement.

3.2 Ridge Nanobeam Cavities

In nanobeam-based on-chip photon sources, the ridge cross-section is critical for maximizing dipole–mode overlap and coupling efficiency ($a$2). The device is segmented into a uniform waveguide, optimally phased Bragg mirrors (period $a$3 satisfying Bragg condition), and an asymmetric cavity of length $a$4. Outcoupling is maximized by asymmetric DBR design (mirror hole count $a$5), and sidewall/position tolerances are numerically robust up to $a$640nm (Gür et al., 2022).

4. Nonlinear and Active Ridge Waveguide Devices

4.1 Second-Harmonic Generation (SHG) in Ridge Geometries

Lithium niobate (LN) and KTP ridge waveguides are crucial for efficient $a$7 processes. SHG performance depends on:

• Cross-sectional size: SHG efficiency $a$8, with effective index splitting quantified via full-vectorial solvers;
• Quasi-phase matching: Periodic poling period $a$9 set from $b$0, with sub-micron accuracy required in fine ridges (Devaux et al., 2015, Boutou et al., 2018);
• Overlap integrals: Fundamental and harmonic mode overlaps, with >98% possible for $b$1 square cross-sections (Devaux et al., 2015).

Cross-sectional uniformity directly affects bandwidth (acceptance $b$2), and fine sidewall/roughness control yields SHG efficiencies close to the theoretical limit bounded only by Fresnel losses (Boutou et al., 2018).

4.2 Superconducting and Metamaterial Ridge Gap Waveguides

Superconducting ridge-gap waveguides employ a central ridge (width $b$3, gap $b$4) between closely spaced plates with periodic pin arrays for mode control, supporting nearly ideal TEM propagation with impedance determined by $b$5 and phase velocity tunable via kinetic inductance (Banys et al., 2021).

Metamaterial ridge waveguides—stacks of alternating Si and SiO$b$6—enable $b$7-dependent control over $b$8, propagation length $b$9, and mode area ($(x/a)^2 + (y/b)^2 = 1$0), with significant trade-offs between confinement and loss governed by fill factor, width, and height. EMT provides accurate, rapid design-space exploration (Sifat et al., 2016).

5. Radiative, Leaky, and Antenna Ridge Structures

5.1 Ridge-Based Grating and Leaky-Wave Antennas

Ridge geometry is fundamental to the performance of millimeter-wave and optical radiators:

• Double-ridged horns: Tapered $(x/a)^2 + (y/b)^2 = 1$1 and a power-law shrink of gap $(x/a)^2 + (y/b)^2 = 1$2 ensure monotonic mode cut-offs and eliminate trapped modes, supporting 10:1 bandwidth with constant beamwidth (Morgan et al., 2015).
• Phased array antennas: Ridge-waveguide gratings achieve sub-degree beam divergences ($(x/a)^2 + (y/b)^2 = 1$3 for $(x/a)^2 + (y/b)^2 = 1$4mm) and sensitivities $(x/a)^2 + (y/b)^2 = 1$5, dictated by grating period $(x/a)^2 + (y/b)^2 = 1$6, coupling coefficient $(x/a)^2 + (y/b)^2 = 1$7, and aperture geometry (Xu et al., 2024).
• Sinusoidal ridges in SIW LWAs: Modulation of amplitude $(x/a)^2 + (y/b)^2 = 1$8 and width $(x/a)^2 + (y/b)^2 = 1$9 tailors leakage rate $w=2a$0 and phase $w=2a$1, enabling control of sidelobe level (<–30 dB) while maintaining beam direction (Mohammad-Ali-Nezhad et al., 2018).

Designers exploit ridge periodicity and asymmetry to achieve field profiles optimal for desired radiative characteristics, minimize cross-polarization, and tune beam steering with high fidelity.

6. Performance Metrics and Geometry–Function Relationships

The table below collates the most salient geometry–performance relationships across representative technologies:

Geometry Parameter	Functional Impact	Example References
$w=2a$2, $w=2a$3	Confinement, $w=2a$4, single-mode cutoff	(Pan et al., 2011, Devaux et al., 2015)
Apex curvature $w=2a$5	SPP loss, confinement, FoM	(Pan et al., 2012)
Period $w=2a$6 (DBR, gratings)	Spectral/angle selectivity, bandwidth	(Gür et al., 2022, Xu et al., 2024)
Ridge width modulation (e.g., $w=2a$7, $w=2a$8)	Stop-band/dispersion control	(Banys et al., 2021)
Ridge amplitude ($w=2a$9)	LWA leakage rate $h=b$0	(Mohammad-Ali-Nezhad et al., 2018)
Fill factor $h=b$1 (metamaterials)	$h=b$2, $h=b$3, $h=b$4	(Sifat et al., 2016)

Performance maximization (e.g., SHG efficiency, SPP FoM, antenna directivity, cavity $h=b$5) requires trade-offs among confinement, loss, and fabrication tolerance. Small cross-sectional areas yield greater intensity and nonlinear conversion, but with heightened mode sensitivity and tighter poling or patterning requirements (Devaux et al., 2015, Gür et al., 2022, Boutou et al., 2018).

7. Fabrication Considerations and Robustness

Ridge waveguide efficacy is strongly modulated by fabrication precision:

• Nanometer-level accuracy in period ($h=b$6, $h=b$7), cavity length ($h=b$8), and etch depth ($h=b$9, $\kappa = b/a^2$0) is required for phase-matching, high $\kappa = b/a^2$1, and efficient coupling (Xu et al., 2024, Gür et al., 2022).
• Material and sidewall roughness (e.g., $\kappa = b/a^2$25 nm RMS in KTP) governs scattering loss and spectral acceptance (Boutou et al., 2018).
• Symmetry and tolerance to position (e.g., $\kappa = b/a^2$3 QD misalignment limiting $\kappa = b/a^2$4 drop to 10%) are documented (Gür et al., 2022).

Emergent design rules advocate for minimizing the cross-sectional area consistent with robust mode coupling, employing smooth structural modulations, and leveraging high-index contrast or metallic surfaces with precise pattern control to optimize ridge waveguide performance across modalities.

---

References: (Pan et al., 2011, Pan et al., 2012, Morgan et al., 2015, Devaux et al., 2015, Sifat et al., 2016, Mohammad-Ali-Nezhad et al., 2018, Boutou et al., 2018, Bezus et al., 2018, Banys et al., 2021, Gür et al., 2022, Xu et al., 2024).