On-chip optical isolators
Abstract
Optical isolation suppresses back-reflections that destabilize lasers, produce parasitic interference, and increase crosstalk in photonic systems. Although free-space and fibre-optic isolators are mature, no broadly adopted on-chip technology yet combines strong isolation, low insertion loss, broad bandwidth, compact footprint, and scalable fabrication. The difficulty is twofold: Lorentz reciprocity must be broken, and every relevant backward spatial, spectral, and polarization channel must be suppressed. This Review develops a common framework for assessing proposed solutions. We first distinguish pairwise nonreciprocity from intensity-asymmetric transmission and from true isolation using the constitutive relations and scattering matrix. We then compare three physical routes to an integrated nonreciprocal response: magneto-optical bias, externally driven space-time modulation, and nonlinear optical processes. Topological and non-Hermitian systems are treated as cross-cutting design frameworks that can route, protect, or enhance an underlying nonreciprocal response, but do not by themselves break reciprocity. Finally, we examine fundamental constraints and practical trade-offs among isolation, insertion loss, bandwidth, footprint, modal coverage, drive requirements, and fabrication compatibility. The resulting mechanism-based perspective connects reciprocity-breaking physics to the requirements of scalable photonic integrated circuits.
Related
- High-speed, compact and low-crosstalk lithium niobate on insulator 2 × 2 optical switch for on-chip integration
- Programmable generation of optical skyrmions on a silicon photonic chip
- Nanoscopic Multiplexing Optical Data Storage via Chip Fabrication
- Research and simulation of analytical polarization control enabled by optical computing on an integrated photonics chip
- Spectro-temporal shaping of broadband optical wavepackets via programmable on-chip photonics
BibTeX
@misc{mikhail2026onchip,
title = {{On-chip optical isolators}},
author = {Abanoub Mikhail and Nikolay Solodovchenko and Sen Xu and Fan Zhang and Maksim Tumashov and Mingzhao Song and Sergei Makarov and Andrey Bogdanov},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.17862},
eprint = {2610.17862},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.17862},
}
RIS
TY - UNPB
TI - On-chip optical isolators
AU - Abanoub Mikhail
AU - Nikolay Solodovchenko
AU - Sen Xu
AU - Fan Zhang
AU - Maksim Tumashov
AU - Mingzhao Song
AU - Sergei Makarov
AU - Andrey Bogdanov
PY - 2026
DA - 2026/10/13
PB - arXiv
DO - 10.48550/arxiv.2610.17862
UR - https://arxiv.org/abs/2610.17862
AB - Optical isolation suppresses back-reflections that destabilize lasers, produce parasitic interference, and increase crosstalk in photonic systems. Although free-space and fibre-optic isolators are mature, no broadly adopted on-chip technology yet combines strong isolation, low insertion loss, broad bandwidth, compact footprint, and scalable fabrication. The difficulty is twofold: Lorentz reciprocity must be broken, and every relevant backward spatial, spectral, and polarization channel must be suppressed. This Review develops a common framework for assessing proposed solutions. We first distinguish pairwise nonreciprocity from intensity-asymmetric transmission and from true isolation using the constitutive relations and scattering matrix. We then compare three physical routes to an integrated nonreciprocal response: magneto-optical bias, externally driven space-time modulation, and nonlinear optical processes. Topological and non-Hermitian systems are treated as cross-cutting design frameworks that can route, protect, or enhance an underlying nonreciprocal response, but do not by themselves break reciprocity. Finally, we examine fundamental constraints and practical trade-offs among isolation, insertion loss, bandwidth, footprint, modal coverage, drive requirements, and fabrication compatibility. The resulting mechanism-based perspective connects reciprocity-breaking physics to the requirements of scalable photonic integrated circuits.
ER -