Skip navigation
Please use this identifier to cite or link to this item:
Title: Strategies for predicting, understanding, and controlling colloidal crystallization
Authors: Reinhart, Wesley F
Advisors: Panagiotopoulos, Athanassios Z
Contributors: Chemical and Biological Engineering Department
Keywords: colloid
machine learning
Subjects: Chemical engineering
Condensed matter physics
Materials Science
Issue Date: 2019
Publisher: Princeton, NJ : Princeton University
Abstract: Colloidal crystals continue to draw attention for their applications in photonics, energy conversion, and biosensing, as well as for their usefulness in elucidating questions in fundamental physics and materials science. However, the wide range of applications for these particles is made possible by the equally wide range of possible self-assembled structures, which may be amorphous or crystalline, close-packed or open, and may include defects with varying morphology and concentration. Experiments, simulations, and theoretical treatments have all shown that this rich phase behavior is enabled by relatively weak interactions between the particles, which leads to small differences in free energy between competing equilibrium and non-equilibrium structures, presenting unique challenges in the fabrication and analysis of the resulting materials. This dissertation explores two complementary topics which both support the goal of engineering advanced materials through directed assembly of colloidal crystals. In the first half of this work, I explore the crystallization of triblock Janus colloids through a variety of simulation methods. I begin by calculating the equilibrium thermodynamic phase diagrams for these particles under a wide variety of conditions, quantifying the driving forces which stabilize each of the competing crystal polymorphs. I next evaluate the crystallization kinetics of several of these solid phases, finding that the different assembly mechanisms of each crystal lattice lead to surprising behaviors. Finally, I demonstrate how a patterned substrate biases the morphology of a crystal film toward large single crystals of a particularly useful polymorph. In the second half, I develop a novel framework for crystal characterization which infers relationships between observed structures automatically rather than relying on pre-programmed templates. I begin by establishing the mathematical basis for this method, showing how my Neighborhood Graph Analysis (NGA) can be used to identify crystal lattices from particle tracking data. Next I extend the framework to binary systems, demonstrating the first-ever automatic characterization of binary superlattices. Finally, I describe a streamlined version of the algorithm which reduces the computational cost by several orders of magnitude, allowing it to be implemented into a fully automated characterization and visualization workflow which is used to investigate a wide variety of crystal lattices.
Alternate format: The Mudd Manuscript Library retains one bound copy of each dissertation. Search for these copies in the library's main catalog:
Type of Material: Academic dissertations (Ph.D.)
Language: en
Appears in Collections:Chemical and Biological Engineering

Files in This Item:
File Description SizeFormat 
Reinhart_princeton_0181D_12868.pdf34.04 MBAdobe PDFView/Download

Items in Dataspace are protected by copyright, with all rights reserved, unless otherwise indicated.