光放大器
放大器
光子学
光电子学
带宽(计算)
宽带
光学
材料科学
净收益
光子集成电路
物理
电信
计算机科学
CMOS芯片
激光器
作者
Nikolai Kuznetsov,Alberto Nardi,Johann Riemensberger,Alisa Davydova,Mikhail Churaev,Paul Seidler,Tobias J. Kippenberg
出处
期刊:Nature
[Nature Portfolio]
日期:2025-03-12
被引量:1
标识
DOI:10.1038/s41586-025-08666-z
摘要
Abstract Optical amplification, crucial for modern communication, primarily relies on erbium-doped fibre amplifiers (EDFAs) 1,2 . Yet, EDFAs only cover a portion of the low-loss spectrum of optical fibres. This has motivated the development of amplifiers operating beyond the erbium gain window. Pioneering work on optical parametric amplifiers (OPAs) 3,4 using intrinsic third-order optical nonlinearity has led to demonstrations of increased channel capacity. OPAs offer high gain, can reach the 3-dB quantum limit for phase-preserving amplifiers and exhibit unidirectional operation. However, power requirements for highly nonlinear fibres 3,5–8 or bulk waveguides 9,10 have impeded their adoption. By contrast, OPAs based on integrated photonic circuits offer the advantages of substantially increased mode confinement and optical nonlinearity but have been limited in bandwidth 11,12 . We overcome this challenge by using low-loss gallium phosphide-on-silicon dioxide 13–15 photonic integrated circuits (PICs) and attain up to 35 dB of parametric gain with waveguides only a few centimetres long in a compact footprint of 0.25 square millimetres. Fibre-to-fibre net gain exceeding 10 dB across an ultra-broad bandwidth of approximately 140 nm (that is, 17 THz) is achieved, with a threefold increase in the gain window compared with C-band EDFAs. We further demonstrate a high dynamic range for input signals, spanning six orders of magnitude, while maintaining a low noise figure. We exploit these performance characteristics to amplify coherent communication signals. This marks, to our knowledge, the first ultra-broadband, high-gain, continuous-wave amplification in a photonic chip, opening up new capabilities for next-generation integrated photonics.
科研通智能强力驱动
Strongly Powered by AbleSci AI