Study On Micro/nano Optical Device

Nanophotonics is regarded as an excellent method to operate light, but conventional designs of optical waveguide gradually expose its limitation, which constrain our capacity to operate lights. The designers of conventional waveguide tend to acquire an initial structure according to analytical theory with following complicated and heavy workload of optimization. Since the analytical theory are limited to special structures, this style of waveguide-design only exploits a small fraction of the infinite structure space. On the contrary, with the prominent advancement of nano-fabrication technology, which endows us with the fabrication of sub-wavelength structure, an arbitrary structure can be realized. Aiming to narrow the gap between conventional design style and advanced fabrication technology, the inverse design method is proposed by researchers.Combining with mathematics optimization, the inverse design method utilizes electromagnetic computational theory to acquire a structure which satisfied objective function of the waveguide among infinite structure space. Conversely, the infinite structure space will bring us with original method to operate light. Additionally, the inverse designed waveguides are equipped with the unparalleled advantage in optics electronic integrated circuit, by its arbitrary shape and footprint theoretically. In addition to design of the passive and linear optical waveguide, the inspiration of compute structure according to objective function can apply to each field of nanophotonics, which is an ideal solution for optics electronic integrated circuit.In this thesis, we investigate the development of the inverse design method in first chapter, followed with the summarization and analysis about the mathematical optimization algorithm and computational electromagnetic involved in inverse design.Furthermore, optimization and extension of existing inverse design computation platform are illustrated. More specifically, the extended definition of initial structure in inverse design is achieved, by which we can set arbitrary shape in the optimization.The definition of objective functionality of the device is critical in the inverse deign,since we theoretical demonstrated the method to calculate the mode field using finite-difference frequency-domain(FDFD). The numerical expression of this method is also included in this thesis.Two essential device power splitter and T-junction, were investigated based on the inverse design method.Firstly, we demonstrate a broadband 1×3 power splitter based on inverse design,with a compact footprint of 2.8×2.8?m2, and a power ratio of 25%:50%:25% at 1550 nm . The insertion losses and ratio variations are less than 0.49dB and 4.4% in wavelength range from 1500nm to 1600nm, respectively. Further investigations show that the separated structure can be removed without degrading the performance of the splitterSecondly, a SOI T-junction with an ultra-compact footprint of 2.8×2.8?m2 is proposed. Simulation results show that the insertion loss is less than 1dB with nearly identical power ratio in wavelength range from 1534 nm to1576 nm , which fluctuates in the range from -49.06% to 50.43% for the upper branch. The calculated mode overlap ratios ? between field of output ports and the fundamental TE mode in input port is higher than 96% during wavelength range from 1520nm to 1580nm. We also fabricated the T-junction and the experimental results conform with our simulation.