# Trimming and ultra-wide bandwidth expansion of silicon frequency comb   spectra with self-adaptive boundary waveguides

**Authors:** Jianhao Zhang, Vincent Pelgrin, Carlos Alonso-Ramos, Laurent Vivien,, Sailing He, and Eric Cassan

arXiv: 1907.11774 · 2019-07-30

## TL;DR

This paper introduces a novel dispersion engineering method using self-adaptive boundary waveguides to significantly expand and trim silicon frequency comb spectra, enhancing bandwidth and phase matching for nonlinear applications.

## Contribution

It presents a new approach to dispersion engineering with self-adaptive boundary waveguides, enabling ultra-wide bandwidth frequency combs in silicon photonics.

## Key findings

- Demonstrated theoretical bandwidth improvements over existing methods
- Enabled low-anomalous dispersion across large wavelength ranges
- Applicable to high-index-contrast platforms for nonlinear applications

## Abstract

Dispersion engineering is among the most important steps towards a promising optical frequency comb. We propose a new and general approach to trim frequency combs using a self-adaptive boundary of the optical mode at different wavelengths in a sub-wavelength structured waveguide. The feasibility of ultra-wide bandwidth dispersion engineering comes from the fact that light at different wavelengths automatically self-adapts to slightly different effective spatial spans determined by the effective indices of the mode. Using this self-adaptive variation on the confinement, we open up the window of low-anomalous dispersion in a large wavelength range, and theoretically demonstrate frequency combs with improved bandwidths with respect to the state-of-art in several different waveguide configurations considered, for a matter of illustration, in the silicon photonic platform. This strategy opens up a new design space for trimming the spectrum of frequency combs using high-index-contrast platforms and provides benefit to various versatile nonlinear applications in which the manipulation of energy spacing and phase matching are pivotal.

---
Source: https://tomesphere.com/paper/1907.11774