UC San Diego

It is important to note that current LIDAR systems rely on rotating the device to scan the field of view in one dimension. Although mechanical rotations allow for a full 360 degrees scan, it requires bulky apparatus that are not so robust to vibrations and harsh environments. Thus, the need for compact, low cost and small-scale solid-state LIDAR is evermore increasing. Silicon photonics seems to be a viable solution for the future of on-chip communication. The advancements in silicon photonics, and CMOS compatible foundries, allow for the fabrication of the next generation of miniaturized optical components and photonic integrated circuits i.e., phased arrays. Silicon photonics has been at the forefront of optical communications and the next generation of optical transceivers. As a result, there has been also some development in the optical phased arrays. Phased arrays are antenna-based devices whereby controlling the phase and amplitude of each individual antenna and due to the interference between them, one can generate an arbitrary emitted pattern. Radio frequency based phased arrays have long been in use for decades and are a big part of data communication. They are considered the "de facto" way to do solid-state electronic beam steering for many RADAR systems. However, devices used for the optical phased arrays offer high pointing accuracy and resolution in a smaller size compared to the radio frequencies due to the much shorter wavelength of light. Despite its immense potentials optical phased arrays have yet to become a commercial product. Part of the reason for that has been the lack of an ideal platform. Existing optical phased arrays are mostly based on silicon and stoichiometric silicon nitride. Although they can be each be useful for specific applications however, they have some drawbacks. First, stoichiometric silicon nitride due to its lower index and thermo-optic coefficient at the telecom wavelengths, it lacks an efficient phase tuning mechanism. Additionally, the waveguiding structure needs to be larger to have a more confined optical mode which makes the footprint and scaling in the system an issue. On another note, silicon based optical phased arrays are not ideal for high-power operation. Due to the losses in the system and the presence of grating lobes, the system requires a greater amount of input power to achieve a given output power. However, this also increases the chance of damage at the input (nonlinear optical losses), where the power is the highest. This is due to the high two/multi-photon absorption coefficient present in silicon. As such, we are proposing a new platform with our high index plasma-enhanced chemical vapor deposited (PECVD) silicon-rich silicon nitride where the positive attributes of both silicon and stoichiometric silicon nitride are combined. The objective of this thesis is to provide a systematic approach towards the development of a novel CMOS compatible platform (high index silicon-rich silicon nitride) which can be an ideal alternative to the existing platform used for OPAs utilizing the thermo-optic effect. We provide a systematic approach towards the development and the enhancement of PECVD deposited silicon nitride's thermo-optic coefficient as a function the deposition parameters and the silicon content. The ratio of the Si/N is adjusted by changing the ratio of the precursor gases used during the deposition (SiH4/N2). We achieve a wide range of linear refractive indices (1.92-3.1 measured at λ = 1550 nm). We show that highly silicon-rich silicon nitride films with index n>3 possess a much larger transparency window extending to the telecom wavelengths while maintaining low loss optical waveguiding in the C-band. We further show that thermo-optic coefficient of our highest index nitride film with n = 3.1 is (1.65 x 10-4 K-1) which is very close to that of crystalline silicon. We then utilized this high index film with a high thermo-optic coefficient to design, fabricate, and characterize photonic integrated circuits that could potentially be used for a phased array system. The cascaded and compact splitters: y-branch and multi-mode interference coupler for splitting the laser beam onto the chip, multimode interferometric switch for amplitude control, thermo-optic phase shifter for phase tuning the antennas, compact bends in single and multi-mode waveguides implemented in the switch and the phase shifter, and the beta mismatched closely spaced waveguides for creating a phase difference between them to avoid coupling in the near-field (even when their evanescent fields overlap). Finally, we were able to fully integrate and experimentally demonstrate an SRN based end-fire facet optical phased array with 16 elements at λ/2 spacing (775 nm) that has a spot size of about 6 degrees at boresight (0 degrees). The wide steering range is close to a full 120 degrees with one dimensional field of view. The high beam quality in our system is due to a) our compact and efficient phase shifter design which allow for localized heating with negligible crosstalk, b) along with the phase mismatched for the closely spaced waveguides antennas. Such system can be scaled and used for the next generation of LiDAR applications.