[Paper Review] Directional Quasi-Phase Matching in Curved Waveguides

Rolf T. Horn, Gregor Weihs|arXiv (Cornell University)|Aug 12, 2010

Photorefractive and Nonlinear Optics3 citations

TL;DR

This paper proposes directional quasi-phase matching (DQPM) in curved waveguides to achieve phase-matching in semiconductors like AlGaAs that lack birefringence or periodic poling. By exploiting the tensor nature of the second-order nonlinearity, waveguide curvature modulates the effective nonlinearity's sign and magnitude, enabling efficient second-harmonic generation and SPDC with potential for high conversion efficiency in integrated photonic devices.

ABSTRACT

In materials that do not allow birefringent phase-matching or periodic poling we propose to use waveguides to exploit the tensor structure of the second order nonlinearity for quasi-phase matching of nonlinear interactions. In particular, we concentrate on curved waveguides in which the interplay between the propagation direction, electric field polarizations and the nonlinearity can change the strength and sign of the nonlinear interaction periodically to achieve quasi-phase matching.

Motivation & Objective

• Address the challenge of phase-matching in semiconductors like AlGaAs that lack birefringent phase-matching or periodic poling.
• Overcome the limitation of traditional quasi-phase-matching, which only reverses the sign of the nonlinearity periodically.
• Explore a new method—directional quasi-phase-matching (DQPM)—that uses waveguide curvature to modulate the effective nonlinearity via orientation-dependent tensor coupling.
• Enable efficient second-harmonic generation (SHG) and spontaneous parametric down-conversion (SPDC) in integrated, compact photonic devices.
• Assess feasibility, loss mechanisms, and design constraints for fabricating DQPM waveguides with tight bends and high index contrast.

Proposed method

• Model second-harmonic generation (SHG) in curved waveguides using the slowly varying amplitude approximation and transverse field polarization assumptions.
• Account for the tensor nature of the second-order nonlinearity $\chi^{(2)}_{ijk}$ (or $d_{ij}$), where coupling strength depends on the relative orientation of field polarizations and propagation direction.
• Use a co-moving coordinate system aligned with the waveguide’s tangent vector $\mathbf{ds}$ to describe how curvature alters the effective nonlinearity over the waveguide path.
• Apply the formalism to AlGaAs, which belongs to the $\bar{4}3m$ point group with non-zero $d_{14} = d_{25} = d_{36}$, requiring field components along all three crystal axes for coupling.
• Simulate DQPM structures using finite-difference time-domain (FDTD) methods, including circular and perturbative designs with 13 unit cells.
• Analyze modal behavior, bend losses, and field polarization alignment to assess feasibility and optimize waveguide geometry.

Experimental results

Research questions

• RQ1 Can waveguide curvature be used to periodically modulate the effective nonlinearity for quasi-phase matching in materials without periodic poling?
• RQ2 How does the orientation of the waveguide’s propagation direction relative to crystal axes affect the strength and sign of the nonlinear coupling?
• RQ3 What are the trade-offs between conversion efficiency, bend loss, and modal fidelity in curved DQPM waveguides?
• RQ4 How do fabrication challenges—particularly sidewall roughness and mode mismatch—impact the performance of DQPM devices?
• RQ5 Can DQPM be extended to non-circular resonators to enhance field intensity and phase-matching efficiency?

Key findings

• Directional quasi-phase matching (DQPM) enables phase-matched second-harmonic generation (SHG) in AlGaAs waveguides by exploiting curvature-induced modulation of the effective nonlinearity.
• In curved waveguides, the effective nonlinearity periodically changes sign and magnitude due to the rotation of the propagation direction relative to the crystal axes, achieving quasi-phase matching without periodic poling.
• FDTD simulations show that light is guided through a circular DQPM structure with negligible loss and polarization orthogonal to the local propagation direction, confirming the mechanism’s viability.
• The maximum curvature of the field path exceeds the waveguide’s curvature, suggesting that optimized designs could achieve higher conversion efficiency than simple circular bends.
• Bend losses and intermodal coupling are significant challenges, especially in high-index-contrast waveguides with tight bends, requiring precise fabrication and high aspect ratios.
• Sidewall roughness is expected to be the dominant loss source due to high index contrast and strong mode confinement, necessitating advanced fabrication control for practical implementation.