Compact, Reconfigurable Optical Delay Line on a Bent Silica Fiber

TL;DR

This paper presents a mechanically reconfigurable optical delay line using a bent silica fiber that creates a tunable SNAP microresonator, enabling continuous delay adjustment with low loss over a broad bandwidth in a compact form.

Contribution

It introduces a novel, mechanically tunable SNAP microresonator in silica fiber for reconfigurable optical delay lines, combining low dispersion, broadband operation, and compactness.

Findings

Achieved continuous delay tuning from 2 ns to 0.5 ns

Demonstrated operation over a 10 GHz bandwidth

Maintained low insertion loss below 6 dB

Abstract

Tunable optical delay lines that simultaneously offer nanosecond-scale delay, broadband operation, low dispersion, and compact footprint remain challenging to realize with conventional integrated photonic platforms. Here we demonstrate a mechanically reconfigurable slow-light delay line based on a surface nanoscale axial photonics (SNAP) microresonator dynamically induced by controlled bending of a silica optical fiber. A localized nanoscale cutoff-wavelength dip, introduced by CO2-laser annealing, provides a reflective boundary, while fiber bending generates a smooth axial potential whose shape is continuously tunable via loop curvature. By adjusting the bending radius, the induced SNAP microresonator evolves from a nearly linear to an approximately semiparabolic axial profile, enabling a controlled transition from dispersive to nearly dispersionless delay. Using a transverse microfiber coupler operated at the impedance-matching condition, we experimentally demonstrate continuous delay tuning from 2 ns to 0.5 ns within a 10 GHz bandwidth in an approximately 2 mm long fiber segment, with insertion loss below 6 dB. The results confirm that mechanically induced SNAP microresonators provide a compact, robust, and reconfigurable platform for dispersion-engineered optical delay lines, with direct relevance to photonic beamforming, frequency-comb stabilization, and neuromorphic photonic signal processing.