The objective of this research is to help bring closer to reality the inevitable merger of optics and electronics, or in material terms, the union of silicon and III V materials on a platform for the next generation of information technology. Specifically, it involves the use of some of the more subtle aspects of guided wave optics to realize spatial control of information-bearing optical beams in crystals.
Intellectual Merit
The intellectual merit of the proposed research is to extend the concept of adiabatic transitions to the optics and communications fields. For years, adiabatic transitions have played a key role in physics and, more recently, chemistry. If successful, the proposed work could have a significant impact on the U.S. optoelectronics industry, contributing to both employment and well-being.
Broader Impacts
Carrying out the proposed research requires the training of graduate students in diverse areas such as electromagnetic modal theory and semiconductor device fabrication. Graduate students engaged in this research will be exposed to ahead-of-state-of-the-art theoretical and fabrication tools. Undergraduate students will be incorporated in the project through Caltech's Summer Undergraduate Research Fellowship program. Results and discoveries from this project will be shared with the scientific and technological communities through journal publications, conference presentations, and seminars.
Hybrid Si/III-V devices are a promising technology to unite efficient electronics fabricated in silicon and efficient optics fabricated in III-V. Developing this platform could lead to faster computer chips and fiber optic networks that simultaneously consume less power than existing technology. This project investigated improving the efficiency of active hybrid Si/III-V optoelectronic devices by engineering the properties of the optical mode at different locations of a photonic integrated circuit. In an optical device like a modulator, amplifier, or detector, optical energy should be mostly in the active material. Where light is traveling between devices, optical energy should be in a low-loss silicon waveguide. This project developed the concept of Adiabatic Modal Control (AMC) to "shuttle" power back and forth between the silicon and III-V materials on a wafer-bonded Si/III-V platform. AMC can improve device efficiency and reduce device footprints, both important metrics for densely-integrated photonic circuits. This project improved AMC theory, including the role of silicon and III-V in different devices which may benefit from using AMC. Highlights of fabricated devices include: Demonstration of hybrid Si/III-V lasers with a threshold current density of 350A/cm2, which is very low for a laser with 5 quantum wells. Making a low-threshold laser was the primary goal of this project. The hybrid Si/III-V lasers have a white-noise linewidth of 18kHz, which is a record for semiconductor lasers tested in air. We directly observe the "shuttling" of the optical mode from a passive silicon waveguide to an active III-V waveguide in order to increase device efficiency. Direct wafer bonding of silicon and III-V wafers has improved to near-100% yield. The hybrid Si/III-V lasers developed under this project have low threshold current densities and superior phase noise properties when compared to other semiconductor lasers. These properties may enable dense integration of the next-generation coherent communication networks which will be necessary to continue increasing data rates. For example, the linewidth of 18kHz demonstrated here would satisfy the requirements for a 16-QAM coherent communication scheme with a bit-error rate better than 10-4 at 40Gbps. This on-chip laser source that couples easily to silicon waveguides and has low threshold and low phase noise enables a new generation of integrated sensors. Currently, fiber lasers are used to interrogate integrated ultra-high-Q sensors because of their stability. Integrating a stable laser source onto a silicon chip could drive down the cost of these sensors considerably and may cause these sensors to become ubiquitous in daily life. Additionally, this grant has funded a new graduate student to be trained in advanced laser design and semiconductor fabrication, including electron-beam lithography, photolithography, reactive ion etching, metal contact deposition, and wafer bonding, and in various characterization methods. This grant partially supported the work of an under-represented minority undergraduate student, mentored by the supported graduate student, to develop and test a method for polishing the edge facets of semiconductor devices for the research group. The undergraduate began the project with no exposure to semiconductor device fabrication, but learned the necessary techniques and the fundamentals of engineering research over the course of the project. He has recently been admitted to graduate school for electrical engineering, a path that likely began partially due to his experience with this project.
