xii, 112 p. : ill. (some col.) / Biological molecular motors exist in an interesting regime of physics where momentum is unimportant and diffusive motion is large. While only exerting small forces, these motors still manage to achieve directed motion and do work. Brownian motors induce directed motion of diffusive particles and are used as models for biological and artificial molecular motors. A flashing ratchet is a Brownian motor that rectifies thermal fluctuations of diffusive particles through the use of a time-dependent, periodic, and asymmetric potential. It has been predicted that a feedback-controlled flashing ratchet has a center of mass speed as much as one order of magnitude larger than the optimal periodically flashing ratchet. We have successfully implemented the first experimental feedback ratchet and observed the predicted order of magnitude increase in velocity. We experimentally compare two feedback algorithms for small particle numbers and find good agreement with Langevin dynamics simulations. We also find that existing algorithms can be improved to be more tolerant to feedback delay times. This experiment was implemented by a scanning line optical trap system. In a bottom-up approach to understanding molecular motors, a synthetic protein-based molecular motor, the "tumbleweed", is being designed and constructed. This design uses three ligand dependent DNA repressor proteins to rectify diffusive motion of the construct along a DNA track. To predict the behavior of this artificial motor one needs to understand the binding and unbinding kinetics of the repressor proteins at a single-molecule level. An assay, similar to tethered particle motions assays, has been developed to measure the unbinding rates of these three DNA repressor proteins. In this assay the repressor is immobilized to a surface in a microchamber. Long DNA with the correct recognition sequence for one of the repressors is attached to a microsphere. As the DNA-microsphere construct diffuses through the microchamber it will sometimes bind to the repressor protein. Using brightfield microscopy and a CCD camera the diffusive motion of the microsphere can be characterized and bound and unbound states can be differentiated. This method is tested for feasibility and shown to have sufficient resolution to measure the unbinding rates of the repressor proteins. / Committee in charge: Dr. Raghu Parthasarathy, Chair; Dr. Heiner Linke, Research Advisor; Dr. Dan Steck; Dr. John Toner; Dr. Brad Nolan