Cellulose nanomaterials (CNMs) have been widely studied for their potential as sustainable fillers in polymer nanocomposites, optical responsiveness in suspensions and thin films, and their orientation-dependent liquid crystalline behavior in suspensions. Cellulose nanocrystals (CNCs) have seen a particular prominence due to their versatility across a breadth of applications. The unique structure of CNCs, represented as nanoscale rods with a slight twist, provides for their self-assembly into liquid crystalline phases when their concentration is increased and can be used to generate iridescent materials with tunable wavelengths. Further, CNCs are often used as fillers in nanocomposites, due to their high single crystal Young's modulus, achieving vast enhancements in stiffness when incorporated above a critical concentration where a percolating network is formed. The breadth of applications for CNCs strongly depend not only on their crystalline structure, but crucially on the interactions between particles. These interactions are well-known, yet a complete understanding to enable the full exploitation of the properties attainable in CNC-based materials is lacking. The principal emphasis of this dissertation lies in further improving our comprehension of the interactions between CNCs across a variety of applications such that their full potential can be achieved. A review of the current research of CNC-based materials is provided to guide the discussion herein. Interparticle interactions are studied in aqueous suspensions of CNCs in evaporating sessile droplets. This system provides a complex interrelationship between mass, heat, and momentum transport which collectively provide a change in the local CNC concentration as a function of time. CNC interactions can be controlled throughout the evaporation process as a result of these local concentration variations. We implement a novel approach using time-resolved polarized light microscopy to characterize the evolution of these particle interactions via the orientation of CNCs as a function of CNC concentration and droplet volume. Ultimately, boundary interactions at the leading edge of the contact line during evaporation was found to drive a cascade of local CNC interactions resulting in alignment post-deposition. Computational analysis evaluated the influence of evaporation-induced shear flow during evaporation. Orientation was found to be independent of the bulk fluid flow, corroborating the importance of interparticle interactions on the ensuing alignment of CNCs. Characterization of an evaporating droplet of initially liquid crystalline suspension of CNCs verified the simulations which predicted that orientation was not coupled with entrainment. Finally, the multiple modes of orientation showed that local control over CNC properties can be realized through governance of the interactions between CNCs. The interactions of CNCs in polymer nanocomposites were also studied for the development of smart materials which can adapt their properties in response to external stimuli. A well-known example of this phenomena is found when CNCs are introduced as fillers in thermoplastic polyurethanes (TPUs) above a critical concentration required to achieve percolation. The interactions between CNCs in the percolating network provide a strong enhancement to the modulus of these materials. However, these materials soften upon exposure to water following the disruption of inter-CNC hydrogen bonding by the diffusing water molecules, as prevailing theories suggest. CNCs simultaneously enhance water transport into hydrophobic matrices. Thus, a complete understanding of the interrelationship between the mass transport and mechanical performance can facilitate the development of humidity sensing or shape memory materials which operate as a result of the interactions between CNCs inside of a polymer matrix. Despite an increase in the equilibrium water uptake with increasing CNC concentration, a decrease in the apparent diffusivity of water within the nanocomposites was observed as a result of swelling of the bulk polymer. Additionally, we developed a modification to the commonly used percolation model to predict the time-dependent evolution of storage modulus during water-induced softening. We found that the solvent mass transport can be directly coupled to the mechanical integrity of the percolating network of CNCs by evaluating the hydrogen bonding state of the network as a function of time. Finally, a novel nanocomposite filler comprised of CNCs and 2,2,6,6- tetramethylpiperidine 1-oxyl (TEMPO) oxidized cellulose nanofibrils (TOCNFs) was prepared through solution casting to improve the mechanical performance of the individual reinforcements alone. The physical interaction length is increased by incorporating CNMs of different length scales resulting in increased tensile strength and elongation. Further, the morphology, evaluated with polarized light microscopy, atomic force microscopy, and simulated with dissipative particle dynamics, revealed the combined fillers exhibit a cooperative enhancement between CNMs. Characterization of the crystallinity through x-ray diffraction confirmed the interactions occur primarily between the crystalline domains of each material. Accordingly, the combination of CNMs resulted in nanocomposite fillers which can be implemented such that the weak interfaces with polymer matrices can be bridged with fillers providing reinforcement over a broader length scale. / Doctor of Philosophy / Cellulose nanocrystals (CNCs) are sustainable and biorenewable nanoparticles derived from cellulose. These materials have been widely studied and are commonly used among a plethora of applications such as in reinforcing fillers in polymer nanocomposites, optically responsive materials that can be used in packaging or anti-counterfeiting technologies, as well as in suspension modifiers for skin care products. These techniques tune the interactions between individual CNCs to modify the behavior of the bulk material. The specific interactions are well-known, yet a complete understanding of the influence of these interactions resulting in the utility of CNC-based materials in various applications is lacking. The principal emphasis of this dissertation lies in further improving our comprehension of the interactions between CNCs across a variety of applications such that their full potential can be achieved. Interactions between CNCs were investigated in three systems comprising of a range of typical use cases for CNC-based materials. The behavior of CNCs was examined in evaporating droplets of aqueous suspensions. These materials exhibited a change in orientation in the final deposit which is dependent on variations in local CNC concentration during drying. These concentration changes describe the relative strength of interactions between CNCs which ultimately influences the final alignment of these materials. Further, these interactions provide a pathway to deposit a controlled orientation of CNCs on a substrate which can be utilized for optically responsive materials or serve as templates for other orientation-dependent materials. CNCs were also incorporated into a thermoplastic polyurethane (TPU) matrix to provide increased stiffness. In these composites, water preferentially interacts with CNCs preventing the nanoparticles from interacting with one another. As water is absorbed, these materials soften as a result of the reduced interactions between CNCs. We investigated the influence of dynamically changing CNC interactions on the mechanical performance of these materials during water absorption and developed an analytical model to describe the observed softening behavior. Finally, CNCs were combined with 2,2,6,6- tetramethylpiperidine 1-oxyl oxidized cellulose nanofibers (TOCNFs) and cast into thin films. The mechanical properties of these differently sized, yet chemically similar, nanoparticles were compared as a function of CNC composition. A cooperative enhancement of the ultimate tensile strength and elongation was observed at low CNC loadings where CNCs and TOCNFs were found to self-organize during casting in a mutually beneficial manner.

L’originalité de ce travail est d’étudier la contribution des nanocelluloses pour lafonctionnalisation des papiers spéciaux. Il y a deux types de nanocellulose, les nanocristauxde cellulose (NCCs) et les microfibrilles de cellulose (MFCs). Il en résulte des propriétésdifférentes à l’état de suspension et à l’état sec. La propriété des MFCs de former un réseaud’enchevêtrement est utilisée pour la dispersion des particules. L’auto-assemblage des NCCsa permis d’élaborer des films iridescents. Ces films ont été considérés comme couchesmodèles puis ensuite mis en oeuvre dans le procédé de fabrication des papiers. Il a été proposéavec succès d’utiliser les MFCs dans le couchage pour réduire la quantité de pigmentsopacifiants pour les papiers minces, et de fabriquer des pigments iridescents pour obtenir despropriétés d’anti-contrefaçon. Ces approches ont été validées à l’échelle laboratoire mais aussipilote. / The original feature of this work is to investigate the contribution of two families ofnanocellulose for their application within specialty papers. It exists two families ofnanocellulose, i.e. Cellulose Nanocrystals (CNCs) and Cellulose Nanofibers (CNFs). It resultsin different properties in suspension and solid states. CNFs with their ability to formentangled network are used as dispersive network for particles. In contrast, the self-assemblyproperties of CNC are used to obtain iridescent films. First, the films based on nanocellulosewere considered as model layers. Then, results were implemented at the industrial scalewithin the papermaking process. It is proposed to use CNF based coating for savingopacifying pigments in lightweight paper, and manufacturing iridescent pigment to impartanti-counterfeiting properties. These sustainable and cost-effective approaches were thenvalidated at pilot scale.

The innovative technology behind production of strong biofilaments involves the process of spinning filaments from nanoparticles extracted from wood. These nanoparticles are called cellulose nanofibrils (CNFs). The spun filaments can have high mechanical properties, rivaling many other plant based materials, and could be an environmentally friendly replacement for many materials in the future such as fabrics and composites. Before mass production might be possible, the optimal dispersion properties must be determined for the intended use, with regard to concentration, method of oxidation (TEMPO-oxidation or carboxymethylation) and pretreatment through sonication and centrifugation. In this bachelor’s thesis attributes of spun filaments were investigated in order to find a correlation between mechanical properties and the effects of concentration, method of oxidation as well as sonication and centrifugation of the dispersions. The mechanical properties were also compared to the fibrils’ ability to entangle and align during flow-focusing. A variety of analytical methods: flow-stop, tensile testing, scanning electron microscopy (SEM) and wide angle X-ray scattering (WAXS) were implemented for the dispersions and filaments. The results from this study show that flow-stop analysis could be used to determine which CNF dispersions are spinnable and which are non-spinnable, along with which spinnable dispersion would yield the strongest filament. It was also concluded that crystallinity of fibrils affects the mechanical properties of filaments and that TCNFs are generally more crystalline than CMCs. Pretreatment through sonication and centrifugation seems to have a negative impact on spinnability and sonication in combination with low concentration seems to lead to non-spinnable conditions. On the other hand, sonicated dispersions seem to yield a greater number of samples without aggregates than non-sonicated ones. Aggregates, however, seem to only affect ultimate stress out of the measured mechanical properties. Furthermore, concentration and viscosity affect spinnability and CMC dispersions seem to yield thicker filaments than TCNF dispersions. However, due to lack of statistically validated data any definitive conclusions could not be drawn.