Nanoparticle diffusometry in hydrogels

Abstract

In order to understand food product functionality such as elastic and flow behavior and mass transport properties, one first has to understand the multi-length-scale structure of the material. The aim of this work is to explore novel methodologies to study and characterize multi-length-scale structures of food hydrogels under static and dynamic conditions. The focus lies on hydrogels comprising polysaccharides, because they show a rich variation in elastic and flow behavior. The largest part of the thesis focuses on the use of nanoparticles (3–30 nm diameter) that are dissolved into the water phase of hydrogels, and whose mobility is reduced due to the presence of the polymer network. This retardation of nanoparticle self-diffusion in hydrogels relative to self-diffusion in neat water can be used to infer structural information about the microstructure of the polymer network. In chapter 2, an in-depth review of existing literature on this method, known as “nanoparticle diffusometry”, is provided, with an emphasis on physical models of self-diffusion in polymer gels and applications in food gels. In that chapter, we distinguish between (1) nanoparticle diffusion in (heterogeneous) polymer gels and (2) nanoparticle diffusion in solutions of (semi)flexible polymers. We adhere to this categorization throughout the rest of the thesis. In chapters 3 and 4 we first describe the design and manufacturing of tailor-made nanoparticles that are functionalized with spectroscopic labels, and the implementation of pulsed-field gradient (PFG) NMR and optical spectroscopy toolboxes for nanoparticle diffusometry. We then use these toolboxes to measure nanoparticle self-diffusion in heterogeneous κ-carrageenan (a polysaccharide) gels. These experiments reveal bimodal nanoparticle self-diffusion (i.e., there are two nanoparticle fractions with different diffusion coefficients) as previously observed in these gels by Loren et al. The results suggest that the sub-micron structure of these gels is heterogeneous with a wide distribution of pore sizes at the sub-micron scale, leading to “sieving” of nanoparticles resulting in the observation of bimodal self-diffusion. This hypothesis is further explored in chapter 5, where besides PFG NMR and optical spectroscopy, Overhauser dynamic nuclear polarization (ODNP)-enhanced NMR spectroscopy is employed. This method can determine the local viscosity of water surrounding the two fractions of particles. It turns out that the particle fraction with the lower apparent diffusion coefficient is in fact trapped in small, nanoscopic interstitials within the gel. The ODNP NMR experiments show that the viscosity of water surrounding the trapped particles is significantly lower than the viscosity within the larger interstitials. Chapter 6 describes a study of nanoparticle diffusion in solutions of poly(ethylene glycol), a flexible polymer with well defined compositions and chain lengths. We use scaling laws to understand the relation between macroviscosity and “microviscosity” as apparent from the nanoparticle diffusivity. We show that the particles probe (near-)macroviscosity only if their size is larger than the size of the PEG polymer coils. Another topic of this thesis is a study of the behavior of food hydrogels under dynamic conditions. To this end we use rheo-MRI velocimetry, which allows us to study the complex shear flow behavior of hydrogels that (per definition) have a yield stress. In chapter 7, we first employ nanoparticle diffusometry to study the sub-micron structure of dispersions of rigid cellulose microfibrils in the presence of carboxymethyl cellulose. Carboxymethyl cellulose is a charged cellulose derivative that succeeds to disperse the aggregation-prone cellulose microfibrils homogeneously at the sub-micron scale. Rheological characterization shows that the resulting dispersions are thixotropic yield-stress fluids. The flow properties of such fluids are well understood, but rheo-MRI experiments show that shear flow of apparently homogeneous cellulose dispersions does not resemble the flow behavior of typical thixotropic yield-stress fluids. We explain the differences by using a fluidity model to show that persistent micron-scale heterogeneity still dominates the flow behavior.