From an Information and Communication Technology (ICT) perspective, the 21st century is characterized by an explosion of requests for communication capabilities, high-performance computing, and cloud storage. Over the last few years, global Internet traffic has been growing exponentially. In this picture, transporting such an amount of data with existing electrical- interconnects and switching technologies will soon reach the "bottleneck" in terms of thermal loading, capacity, latency and power consumption. Optical- interconnects and switch fabrics combined with photonic integrated circuits (PICs) are seen as one of the most promising routes to push such limits. Silicon (Si) photonics is now considered as a reliable photonic integration platform. The beauty of Si Photonics stems from its ability to integrate microelectronics and photonics on a single Si chip utilizing standard CMOS IC technology. An important subset of this area is hetero-integration of III-Vs on Si, where the aim is the make use of III-V materials, with superior optical properties, to provide an efficient optical gain medium to circumvent the fundamental physical limitation of Si, i.e. Si cannot efficiently emit light, yet keeping the capability of light-routing, modulating, detecting and cost advantages of Si.
In a breakthrough development, the investigators' group in UCL have shown that it is possible to grow epitaxially high-performance quantum dot (QD) lasers directly on Si substrates, opening up the possibility to monolithically integrate various types of III-V optoelectronic devices on Si. The pace of research on monolithic III-V/Si integration has then been dramatically accelerated and an increasing number of prestigious research groups including Bowers' group at UCSB and Arakawa's group at Tokyo University, and major Si chip companies, i.e. Intel, are currently devoting considerable programmes in this area. In addition to III-V/Si lasers, monolithic III-V/Si semiconductor optical amplifiers (SOAs) are also attracting significant interest as the key components for next-generation photonic integrated optical- interconnects and switching fabrics, as the application of SOAs is not limited only to compensate for loss and maintain signal levels as the signal propagates throughout a large number of optical components within the PICs, it is also used as a mature gating element for optical switches and has the advantages of ease of control, smaller footprint, low operating voltage, high ON/OFF extinction ratio, and fast transition times of the order of nanoseconds. However, such a III-V/Si SOA has not been developed to date.
Building on the established expertise in monolithic III-V/Si QD lasers at UCL, this project proposal aims to develop the world's first monolithic III-V QD SOA on CMOS-compatible on-axis Si (001) substrates. In contrast to conventional native substrate based SOAs or III-V/Si SOAs using either flip-chip bonding or wafer bonding, the proposed method is fundamentally different, since the III-V SOAs will be integrated on Si by direct epitaxial methods, offering the possibility to achieve high-yield, low-cost and large-scale Si-based PICs, which is expected to be the technology platform to address next-generation optical- interconnect and switching solutions. With further development in Si photonics, i.e., providing the microelectronics world with the ultra-large-scale integration of photonic components, there will be scope to target applications in important areas such as consumer electronics, high-performance computing, medical and sensor solutions, and defence.
This project will benefit from guidance from and joint work with both industrial as well as academic partners and will leverage major UK-based industrial and academic strengths in materials (e.g., CSC, EPSRC NEF) device processing (e.g., EPSRC CSHub, Glasgow) and photonics (e.g., Rockley, Lumentum), who are also well positioned to exploit this research.
|