The accelerating global digitalization has given rise to demanding scenarios such as real-time computation of high-definition video streams and real-time rendering of 3D scenes. Traditional electrically-switched architectures with static resource configurations are constrained by bandwidth bottlenecks and protocol stack overhead, making it difficult to meet the requirements of ultra-low latency and ultra-high throughput. AI-driven all-optical networks, enabled by intelligent traffic prediction and dynamic optical path reconfiguration, provide adaptive and real-time scheduling capabilities, emerging as a key direction for the evolution of all-optical data center networks.

Electro-optic polymers (EOPs), owing to their high-speed modulation capability and CMOS compatibility, have become important materials for constructing hybrid integrated electro-optic waveguides. However, electric field poling in Si?N?/EOP hybrid waveguides to induce the Pockels effect often suffers from low poling efficiency or even dielectric breakdown due to discontinuous electric field distribution and interface roughness. To overcome this bottleneck, this work proposes a multi-step poling strategy that effectively enhances the alignment efficiency of chromophore molecules in the cladding, significantly improves tuning performance, and achieves a high effective electro-optic coefficient of up to 114 pm/V without causing breakdown.

Ultrafast all-optical processing functional devices with high response speed, low power consumption, and easy integration have now achieved significant applications in optical communications, quantum computing, optical sensing, and data processing. Ultrafast light modulation and efficient wavelength conversion can be realized by manipulating light-matter interaction, particularly through nonlinear optical effects. However, the weak nonlinearities of most optical materials, along with the phase mismatch issue, limit their application in signal processing.

Optical filters are extensively employed in wavelength division multiplexing (WDM) systems, microwave photonics, external cavity lasers, and optical differentiators. The implementation of these filters relies on various photonic integrated circuits (PICs), such as a long-period grating (LPG), Bragg grating, grating-assisted directional coupler, Mach-Zehnder interference (MZI), micro-ring resonator (MRR), MRR-MZI, etc. To process flexibly the optical signals in the next-generation reconfigurable WDM net-work, optical filters with bandwidth and wavelength tunability simultaneously are highly desirable and have attracted ever-increasing attention. To enlarge the wavelength and the bandwidth tuning range, the optical filters based on MRR-cascaded, MRR-assisted-MZI, and MZI-assisted-MRR structures have been proposed. Most prior MZI-assisted-MRR filters are realized on mature material platforms such as silicon on insulator (SOI) and silicon nitride platforms.