弯曲分子几何
功率分配器和定向耦合器
宽带
光学
材料科学
波长
集成光学
波分复用
混合耦合器
光电子学
联轴节(管道)
硅
硅光子学
光纤
物理
冶金
复合材料
作者
Ahmed H. El‐Saeed,Alaa Elshazly,Hakim Kobbi,Rafal Magdziak,Guy Lepage,Chiara Marchese,Javad Rahimi Vaskasi,Swetanshu Bipul,Dieter Bode,Marko Ersek Filipcic,Dimitrios Velenis,Maumita Chakrabarti,Peter De Heyn,Peter Verheyen,P. Absil,Filippo Ferraro,Yoojin Ban,Joris Van Campenhout,Wim Bogaerts,Qingzhong Deng
标识
DOI:10.1109/jlt.2024.3407339
摘要
The wavelength-insensitive response of the bent directional coupler (DC) is rigorously studied by means of coupled mode theory analysis. On that basis, we demonstrate an experimental-based coupling model for the bent DC capable of achieving wavelength-insensitive coupling for arbitrary coupling ratios. The measured devices exemplify wavelength-insensitive coupling at key ratios: 0.4, 0.5, 0.6, and 0.7. As a benchmark, we demonstrate a 0.5:0.5 splitter that significantly reduces coupling variation from 0.391 in the traditional DC to just 0.051 over an 80 nm wavelength span. This represents a remarkable 7.67 times reduction in coupling variation. Further, newly-invented low loss bends were used in the proposed design leading to a lossless design with negligible excess loss ( $<0.01$ 0.01 dB). The proposed 0.5:0.5 design is tolerant and shows consistently low cross coupling variation over the 300 mm wafer showcasing a maximum cross coupling variation of 0.112 with a corresponding average cross coupling of 0.531 over 80 nm wavelength range, at the extreme edge of the wafer. Futhermore, we augmented the wafer mapping with a waveguide width fabrication tolerance study, confirming the tolerance of the device with a mere 0.061 maximum coupling variation with a waveguide width deviation of ±20 nm over 80 nm wavelength range. The proposed splitter meets essential requirements for 2 × 2 splitters such as wavelength-insensitivity, support for arbitrary coupling ratios, lossless coupling, fabrication tolerance, and compact footprint, offering a promising application prospect for mass production.
科研通智能强力驱动
Strongly Powered by AbleSci AI