Abstract
In this paper, we have (1) built a micro flow network that mimics micro vessels and visualized the flow inside utilizing both experimental and CFD approaches, (2) developed a model, which considering both the inertial forces caused by flow convection and the diffusion resulting from Brownian motion, to evaluate the entrapment probability of nanoparticles with various particle sizes and channel geometries from micro flow field into nanochannels.. The micro flow network was fabricated by photolithography process and Deep Reactive Ion Etching (DRIE) process. Hundreds nanometer wide nano-gaps located around curved silicon micro-channels were arranged for simulating the geometry of leaky blood vessels in tumors. The nano-gap was about 600 nm in both width and depth, and fabricated by focus ion beam (FIB) process. Finally, the chip was sealed with a Pyrex glass substrate by anodic bonding technique. Micro particle image velocimetry (μPIV) is then employed to visualize the particle-laden flow inside the microfluidic channels. By dual pulse μPIV, the particle images are recorded at two specific moments with continuous illumination. In the experimental results, the spatial resolution of about 2 μm is employed to resolve the near-wall flow field with 50% interrogation spot overlapping by using μPIV. The most significant deviation is about 5% between the experimental and simulation results, which implies excellent agreement and brings out the evidence in utilizing simulation for more detailed study on the flow behaviors of nano- particles. We found that the working fluid passing through the inner side of the curved channel was apparently accelerated. The distortion in velocity profile could potentially lead to the increase of particle concentration, and facilitate the entrapment of particles into the nano-gap. And finally we used Stokes-Einstein equation to estimate the entrapment probability of nanoparticles with various particle sizes and channel geometries.