Process-Structure-Property Relationships of Cellulose Derivative/Thermoplastic Urethane Polymer Composites

Abstract

Cellulose-reinforced thermoplastic composites have garnered significant attention as sustainable alternatives to conventional petroleum-based materials, offering advantages such as biodegradability, mechanical reinforcement, and tunable functionality. These composites hold immense potential for next-generation applications in soft robotics, flexible electronics, and adaptive structures. However, the widespread adoption of such systems remains limited by processing challenges, particularly the difficulty of incorporating cellulose derivatives into thermoplastic matrices without compromising dispersion and thermal stability. Many cellulose-based composites are often fabricated through solution casting using hazardous organic solvents to achieve better dispersion, which sacrifices their scalability and geometric versatility. To fully harness their potential, it is essential to transition these systems toward melt-processable, extrusion-compatible formats. Aligning with additive manufacturing (AM) technologies, particularly fused filament fabrication (FFF), opens additional opportunities for tailored structural design. Achieving this transition requires a fundamental understanding of the structure–process–property relationships that govern performance across different fabrication methods and filler architectures. This dissertation focuses on the development of 3D-printable thermoplastic polyurethane (TPU) composites reinforced with hydroxypropyl cellulose (HPC) or cellulose nanocrystals (CNCs), aiming to systematically investigate how formulation, processing, and filler interactions dictate thermal stability, rheological behavior, mechanical performance, and water-responsive properties. Through a series of three integrated studies, the evolution of these composites from solvent-cast films to fully extrudable, AM-compatible filaments is established. In Chapter 3, an HPC/TPU blend system was fabricated via a traditional solvent-casting approach to establish a baseline understanding of the material's water-triggered optical switching behavior and mechanical adaptability. The blend films exhibited reversible opacity changes upon hydration, attributed to refractive index changes introduced by the swelling and deswelling of HPC phase-separated domains. Dynamic mechanical analysis under submersion conditions revealed modulus softening due to water plasticization, suggesting potential applications in moisture-sensitive optics and soft actuators. Chapter 4 transitions the system in Chapter 3 to a solvent-free, melt-extruded HPC/TPU composite filament designed for compatibility with material extrusion additive manufacturing (MEX). Detailed thermal (TGA) and molecular (SEC, FTIR) analyses confirmed the thermal and chemical stability of both HPC and TPU under extrusion conditions. Rheological testing identified a processable temperature window of 200–230 °C, beyond which TPU exhibited oxidative crosslinking. The composite demonstrated enhanced mechanical strength at 10 wt% HPC loading, and the 3D-printed film retained reversible humidity-responsive optical behavior, establishing a robust structure–property link under melt processing constraints. Chapter 5 presents the development of a 3D-printable cellulose nanocrystal (CNC)/thermoplastic polyurethane (TPU) nanocomposite system with spatially programmable mechanical properties achieved through controlled variation in CNC concentration. To ensure adequate dispersion, CNC/TPU composites were first prepared via solvent casting and then melt-extruded into filaments suitable for single-nozzle extrusion-based 3D printing. Thermal and rheological analyses established a stable processing window between 200–220 °C, within which the composites maintained thermal integrity and favorable flow behavior. Frequency and temperature sweep rheology revealed CNC-induced increases in complex viscosity and enhanced shear-thinning, attributed to the formation of a percolated hydrogen-bonded CNC network that restricted TPU chain mobility. Mechanical characterization demonstrated a clear, concentration-dependent enhancement in stiffness, with tensile and dynamic mechanical analysis (DMA) confirming gradient mechanical behavior across the CNC/TPU series. Post-immersion DMA further revealed fully reversible softening attributed to the temporary disruption of the CNC network in the presence of water, followed by recovery upon drying. This stimulus-responsive behavior, combined with tunable stiffness, underscores the potential of this system for adaptive mechanical applications. The filament-level compositional design introduced in this chapter provides a scalable, material-efficient pathway for fabricating mechanically heterogeneous and environmentally responsive structures, with broad relevance to soft robotics, bioinspired actuators, and multifunctional 3D-printed devices.