Researchers from the Institute of Microelectronics at Peking University have made a significant stride in the development of integrated optical transceivers, which are crucial for data-intensive applications like artificial intelligence and cloud computing. Their work, published in the journal Nature, focuses on the heterogeneous integration of thin-film lithium niobate (TFLN) with active silicon photonics, aiming to create energy-efficient, high-capacity photonic systems.
The team, led by Professor Dapeng Liu, has successfully demonstrated a novel approach to integrate TFLN with a fully functional silicon photonics platform. This is achieved through a process called trench-based die-to-wafer bonding, which allows the introduction of TFLN after the completion of all CMOS-compatible processes for silicon photonics. This sequential integration overcomes the process incompatibilities that have previously limited the direct integration of TFLN with only passive silicon photonics.
The integrated chip features a variety of components, including low-loss fiber interfaces, 56-GHz germanium photodetectors, and 100-GHz TFLN modulators, all interconnected through multilayer metallization. The researchers have achieved efficient inter-layer and inter-material optical coupling, resulting in on-chip optical links with greater than 60 GHz electrical-to-electrical bandwidth. These links support 128-GBaud on-off keying (OOK) and 100-GBaud pulse amplitude modulation 4-level (PAM4) transmission below forward error-correction thresholds.
The practical applications of this research are significant for the energy sector, particularly in data centers and high-performance computing environments. The integrated optical transceivers can facilitate large bandwidth, low power consumption, and high integration density, which are essential for the explosive growth of data-intensive applications. By adopting CMOS processes, the technology also offers the potential for low-cost manufacturing, making it a scalable platform for energy-efficient, high-capacity photonic systems. This advancement could lead to more efficient data transmission and processing, ultimately contributing to the reduction of energy consumption in the digital infrastructure.
This article is based on research available at arXiv.