This dissertation demonstrates the utility of high-resolution microscopy methods to visualize the biomimetic crystallization mechanisms that dictate protein localization and spatial distribution pattern outcomes. Biomineralization is the intricate biochemical process through which living organisms create hierarchically structured organic-inorganic composites that offer crucial biological functionalities like catalysis, storage, and protection. Biomineralization processes have inspired the biomimetic crystallization of protein-based metal-organic framework biocomposites (p@MOFs) to form biohybrid materials for potential applications in biomedicine and biocatalysis. A critical knowledge gap in our understanding of p@MOF structure-function relationships is how the precise spatial localization and distribution patterns of the proteins embedded within the MOF crystals affect their overall properties. Furthermore, the spatiotemporal tracking of protein molecules during the biomimetic crystallization process has yet to be fully elucidated, which is essential for designing synthetic procedures to target specific structures and functions.

Chapter 1 outlines fundamental concepts of classical and non-classical crystallization, focusing on techniques used to decipher the key parameters governing crystal nucleation and growth in biomolecule-directed and biomimetic systems. In Chapter 2, the naturally occurring iron oxide mineral core of the protein horse spleen ferritin (Fn) is leveraged as a contrast agent to directly observe individual proteins during the crystallization pathways and outcomes of p@MOFs using transmission electron microscopy (TEM) methods. Chapter 2 illustrates how the 3D mapping of proteins in the crystallization of p@MOFs using advanced TEM techniques can be utilized to analyze protein spatial distribution and aggregation outcomes quantitatively. In Chapter 3, different electron microscopy modalities such as cryogenic TEM (cryo-TEM), cryo-electron tomography (cryo-ET), and liquid phase TEM (LP-TEM) are explored to understand further how proteins influence both classical and nonclassical crystallization pathways. The collective work in Chapters 2 and 3 demonstrates that adjusting the MOF precursor ratios and protein concentrations alters how proteins are arranged, localized, and aggregated within MOF crystals. Furthermore, this localization is intimately tied to the underlying crystal formation mechanism. This work gives essential insights into the influence proteins have on the crystallization pathways of MOFs, serving as a potential roadmap for designing highly controlled advanced materials capable of improving protein immobilization, performance, and delivery.

Finally, Chapters 4 and 5 describe the effects of incorporating proximal Lewis acids and electric fields in CO2 activation. Chapter 4 outlines how oriented electric fields have been used to investigate their impact on catalytic activity in synthetic and biological systems. Chapter 5 illustrates heterobimetallic complexes with incorporated alkali cations close to a redox-active transition metal center for CO2 activation and reduction. This work examines how positioned proximal cations can be used in synthetic systems to enhance catalytic performance.

Protein-Metal-Organic Frameworks (MOFs) present a diverse array of building blocks for designing high-performance materials with applications spanning numerous industries including drug delivery, catalysis, and gas storage. However, the intricate crystallization mechanisms involving various pathways and intermediates complicate their understanding, stalling the implementation of the materials. Fundamental crystallization studies have been pivotal in advancing both biological and synthetic systems, and MOFs are no exception. This dissertation delves into how proteins exert control over the nucleation and growth of MOFs, influencing final crystal properties such as topology, morphology, and encapsulation efficiency. Advanced microscopy and scattering techniques, such as cryogenic transmission electron microscopy (cryoTEM) and powder x-ray diffraction (PXRD), are employed to conduct these studies. The investigation commences with zeolitic imidazolate framework-8 (ZIF-8) as a model MOF system and bovine serum albumin (BSA) as a model protein. Synthetic protocols are established to explore the impact of BSA concentration and ligand-to-metal ratios on ZIF-8 formation. Notably, the use of cryoTEM allows for the direct observation of MOFs, uncovering the role of transient amorphous phases in ZIF-8 nucleation and growth. Furthermore, in a follow up study, it was discovered how protein folding influences the stability of these amorphous phases, enabling modulation of key physical properties such as crystal size, morphology, and encapsulation efficiency. Moving beyond BSA, subsequent investigations aim to connect nucleation and growth mechanisms to the final enzymatic performance of MOFs using catalytically active proteins, namely glucose oxidase (GOx) and catalase (CAT). The results underscore the significance of protein folding in the initial complexes with MOF precursors, pivotal in designing active materials. To further tailor the extent of enzyme activity, the accessibility of the enzyme to substrates and its diffusion barrier can be controlled through final crystal properties, including size, topology, defects, and porosity. Generalizing these findings, the dissertation introduces two novel, catalytically active proteins into ZIF-8 for the first time, main protease (Mpro) of SARS-CoV-2 and Nanoluciferase (NanoLuc). In summary, this work provides valuable insights into how protein folding influences MOF nucleation and growth mechanisms and, consequently, their final crystal properties. The research contributes to the broader understanding of MOFs as versatile materials and paves the way for high-performance hybrid materials.

Molecular self-assembly is the spontaneous organization of molecules into higher-ordered nano or microstructures. Considering the free energy landscape, self-assembly can be highly pathway dependent, leading to the formation of structures not at the free energy minimum or non-equilibrium. Some non-equilibrium systems arise through chemical reaction-induced assembly. Living organisms rely on reaction-driven non-equilibrium assemblies for growth and adaptability. Thus, to develop biomimetic materials, we will have to gain a better mechanistic understanding of non-equilibrium self-assembly. The focus of this dissertation is on the development and mechanistic analysis of two classes of reaction-driven non-equilibrium systems: polymerization-induced self-assembly and transient or dissipative supramolecular assembly. In the first part of the dissertation, I developed a one pot scalable and reproducible synthesis for polyester-based block copolymer nanostructures coined ring-opening polymerization-induced crystallization-driven self-assembly (ROPI-CDSA). In ROPI-CDSA, mPEG is chain-extended by L-lactide in toluene with the assistance of an organocatalyst forming semicrystalline poly-(L)-lactide-block¬-polyethylene glycol (PLLA-b-PEG) block copolymers. As the polymer chain grows, it becomes increasingly solvophobic, leading to self-assembly. The resulting nanomorphology can be controlled by modifying the polymer concentration or PLLA chain length. Additionally, through catalyst selection, both the onset and rate of self-assembly can be controlled. Fast polymerizations result in a rate of polymerization faster than a rate of self-assembly leading to non-equilibrium assembly. A consequence of this disparity is that through catalyst choice, we can access different morphologies without any variation in molecular structure. Next, in ROPI-CDSA, traditional organocatalytic catalysts are swapped for antibiotics with similar functional groups. The resulting drug-embedded polymer nanoparticles retain antibacterial activity. I postulate that this drug-catalyst strategy can be extended to other examples of polymerization-induced self-assembly for the scalable one pot production of drug-loaded polymer suspensions. A hybrid photoiniferter ROPI-CDSA system is also developed expanding ROPI-CDSA to a variety of biocompatible polyacrylamido-b-polyesters. The second part of the dissertation focuses on the investigation of two transient supramolecular systems. One system is the antibiotic teixobactin which forms sheet-like assemblies. At higher concentrations, these assemblies selectively act upon gram positive bacterial cell walls, whereas at lower concentrations aggregates are either formed transiently or sparingly at the cell surface. The other system is a chemically-fueled redox system, where oxidation of a thiol into a disulfide hydrogelator leads to nanofiber assembly. As the fuel depletes, the disulfide is reduced back to the thiol precursor, leading to disassembly. I study the assembly and disassembly processes sequentially and synchronously using time-resolved cryoEM. I show that a thermodynamically unstable stacked nanorod phase observed transiently in the sequential process is sustained throughout the synchronous process showing that energy input drives non-equilibrium self-assembly. As the systems studied in this dissertation mirror biological processes, further understanding of non-equilibrium molecular self-assembly is key to the development of the next generation of robust, adaptable, and active materials as well as unlocking the origins of life.

In situ transmission electron microscopy enables the study of nanoscale solution state events such as battery electrode deposition and molecular self-assembly. Recent improvements in electron microscope detectors have vastly improved the spatial resolution (pixel count) and temporal resolution (frame rate) of microscopy videos, which has resulted in large, information-rich datasets. The central mission of my PhD has been to use image analysis techniques and algorithms to translate improvements in microscopy signal collection to improvements in materials insight. Towards this, I have developed several novel image analysis algorithms built on deterministic image science methods which extract quantitative information from raw microscopy datasets. I developed an algorithm to quantify materials dynamics captured with in situ microscopy videos and applied this algorithm to quantify the spatiotemporal dynamics of dissipative self-assembly processes (Chapters 1-2). I applied image science to track the presence of an unstable phase captured in time-resolved cryogenic transmission electron microscopy (Chapter 3). And finally, I applied a variety of custom image analysis algorithms to quantify amphiphilic block copolymer morphological transitions such as vesical-to-bilayer transitions, droplet-to-bilayer transitions, and droplet dissolution dynamics (Chapters 4-5). In each case, I demonstrate the power of deterministic image analysis methods for converting raw microscopy datasets into interpretable materials science results.