We developed optical methods to manipulate molecules in a microfluidic environment. Optical tweezers can manipulate micro-spheres in solutions with the gradient force but are not practical for spheres smaller than 500 nm in diameter. Nanotweezers use the evanescent field out of waveguides, slot-waveguides, plasmonic resonances, and photonic crystal resonators. They were able to manipulate objects down to 40 nm. Proteins and many biomolecules are of sizes on the order of a few nanometers, a priori out of reach of these techniques. During my PhD, I developed nanophotonic and nano-optic systems aimed at applying electromagnetic potential wells to bias the motion of molecules against Brownian motion and eventually demonstrated that chemical reaction pathways could also be altered. I showed that photonic crystal resonators are a toolbox for nanoscale assembly enabling trapping, transport, and orientation of nano-objects. I also investigated the heat arising in optofluidic photonic crystals and found it to be higher than previously thought, up to 57 K for 10 mW of power input, which makes such devices incompatible with biological single molecule experiments. I then used electromagnetic fields shaped by waveguides-carbon nanotubes hybrids to trap immunoglobulin of mass down to 160 kDa. Last, I developed the optical manipulation of chemical reactions. I showed that electromagnetic gradient force can transport molecules across reaction barriers along a reaction coordinate demonstrating it experimentally by guiding the adsorption of immunoglobulin proteins onto carbon nanotubes. These techniques are part of a wider evolution that is changing the way we interact with molecules. Although originally dismissed for studying single molecules because of the diffraction limit, nano-optics and nanophotonics are becoming the center of this revolution.