Optofluidic particle manipulation provides a powerful and versatile technological platform for on-chip sensing. Embedded planar nanophotonic devices can shape electromagnetic fields in fluidic channels, allowing for a high level of control over particles. This thesis reports my research contribution to designing optofluidic nanostructures for several different kinds of on-chip particle manipulation that are detailed as below. I have numerically demonstrated plasmonic nanoparticle routers that can guide and route nanospheres in a microfluidic channel. I have analyzed the power flow and the corresponding optical force on the nanosphere, and have derived the Maxwell stress tensor utilized in the finite element analysis solver. I also identified the relationship between the relative refractive index of the nanospheres and the magnitude of the generated optical force. The results suggest a new method for next-generation plasmo-fluidic sensing. I have designed dielectric metalenses with phase profiles that can be coherently controlled. The Mie scattering field from the meta-atoms of the metalens can be tailored dynamically, in which the output Bessel beam sweeps in a range from –1.37° to 1.36°. I have further analyzed particle routing in a continuous flow. I have numerically demonstrated a metalens microfluidic microsphere sorting based on fluorescent color. The sorting originates from the metalens’ ability to focus fluorescent light back onto the target sphere, creating self-induced optical tweezers. Because the embedded metalens doublet eliminates the need for any additional sorting mechanism, the technique can be referred to as FEACS (Fluorescence-Enabled