Prediction and control of organic xerogel microstructure : experiments and modelling

Student thesis: Doctoral Thesis

Abstract

Resorcinol-formaldehyde (RF) gels are a fairly new addition to the variety of porous materials used by current industry. However, a better understanding of their formation processes is crucial for efficient structure tailoring, leading to high-performance materials for a range of applications. The usual RF gel manufacture process involves gelation of an RF sol at elevated temperatures, followed by exchanging water within the structure for another liquid, in order to limit shrinkage during the final drying step. Yet, despite significant research efforts, the RF growth processes are still not fully understood and there is no accepted model describing these processes. Therefore, this study combines experimental and simulation approaches to better understand the gel growth process, and how the resulting structure depends on the growth conditions. The experimentally investigated areas include the influence of sodium carbonate catalyst concentration, processing temperature, solvent exchange and drying methods, as well as the presence of different anions within the reaction solution. Discussion on optimal processing parametersis included, in order to preserve the majority of the porous structure of RF xerogel materials, taking process economics into account, and the diversity of textural properties for obtained materials is examined. In order to model the growth processes in RF gels, and investigate how they impact the structural properties of final materials, a two-dimensional lattice-based computational model, using kinetic Monte Carlo, was developed in this work. The presented model is developed to capture growth from monomeric species present in the initial stages of the gelation composition. Experimentally, gel growth is primarily controlled through catalyst concentration, which determines the density of species that are activated for rapid growth, and solids concentration; the model captures both of these dependencies.Generated cluster structures were analysed for textural properties, such as accessible porosity and accessible surface area, as well as fractal properties, in the form of the correlation dimension and the Hurst exponent. Increasing both solids content and percentage of activated monomers led to an observed increase in complexity of cluster arrangement and tortuosity of pore structure, both reflected in the values of evaluated fractal properties. In order to allow comparison of generated cluster structures with trends observed for experimental samples, gas sorption was modelled here using a lattice gas in a mean field approximation. The observations for model pores with varying dimensions agree with the background theory and the trends observed for the cluster structures were in line with those obtained experimentally. This helps to close the loop from growth processes to textural properties, providing the possibility to tailor materials for specific application.
Date of Award26 Sep 2019
Original languageEnglish
Awarding Institution
  • University Of Strathclyde
SponsorsUniversity of Strathclyde
SupervisorPaul Mulheran (Supervisor) & Ashleigh Fletcher (Supervisor)

Cite this