Addressing and two-dimensional driving of droplets on single-sided continuous optoelectrowetting chip

Digital microfluidics(DMF) represented by electrowetting on dielectrics(EWOD) has a broad application prospect in bioassay, chemical synthesis and drug discovery, but currently EWOD technology faces problems such as wiring bottleneck and complexity of droplet detection circuits. Optoelectrowetting(OEW) addresses the wiring bottleneck issue by using light pattern to control droplets. To date, light patterns have been designed by hand and have only allowed one dimensional manipulation. In this paper, we introduced a method using programmable optical patterns to modulate the electric field distribution on the chip, which realized two dimensional manipulation of droplet on the chip. At the same time, machine learning target detection algorithms YOLOv3 was used to accurately detection of droplets the self-addressing of light patterns.