Sugiyama, Masano2010-01-062010-01-062009-08https://hdl.handle.net/11299/56229University of Minnesota Ph.D. dissertation. August 2009. Major; Chemical Engineering. Advisor: Victor H. Barocas. xiii, 153 pages. Ill. (some col)X-ray diffraction is the most common way to determine protein structure at an atomic level. To determine the protein structure, a high-quality crystal of sufficient size is required. Obtaining such a crystal is difficult due to the multi-parametric phase space that needs to be screened to determine the best conditions for growth of a suitable crystal. In this work two microfluidic protein crystallization techniques have been developed and tested: the continuous-feed crystallization chamber and the phase diagram visualizer. The continuous-feed crystallization chamber (CCC) allows for kinetic path control through the crystallization phase diagram during crystallization. The CCC operates similarly to a continuously stirred tank reactor, where protein, salt, and buffer are fed at desired flow rates and concentrations to maintain desired conditions inside the chamber. A lumped kinetic model was developed, and parameters for heterogeneous nucleation kinetics were determined. Heterogeneous nucleation was found to have faster nucleation kinetics and slower growth kinetics than homogeneous nucleation, as expected. The lumped-model analysis gives a method to quantifying the effect of various crystallization variables by extraction of kinetic parameters. The phase diagram visualizer (PDV) determines the solution phase diagram for protein-precipitant systems in one experiment rather than many lengthy experiments as required for traditional methods. Laminar flow and diffusion in the PDV create significant gradients in concentration, so crystals form in only part of the chamber. By combining observation of the location of the crystal-rich regions with a computer simulation of flow and transport in the chamber, a solution phase diagram is generated. This PDV has been tested for the lysozyme-sodium chloride and lysozyme-soidum nitride system. Modeling results where used to design an improved PDV with grooves. This device has been fabricated and is to be tested in the next phase of experiments. These two microfluidic devices together can be used together to determine and execute an optimized growth strategy for a given protein or a condition change. The PDV will give a general road map of the phase space that will be traveled using the CCC.en-USCrystal GrowthMEMSMicrofluidicsPhase DiagramProtein CrystallizationChemical EngineeringThesis or Dissertation