Computational description of porous material: An aerogel use case

Abstract

Reconstructing the intricate morphology of nanoporous materials, such as aerogels, poses a formidable challenge when one aspires to attain three-dimensional visual representations of their mesoporous (pore sizes between 2 and 50 nm) nanostructure. The array of available microscopic and tomographic instruments encounters formidable hurdles in penetrating the diverse spectrum of aerogel types, hindering the comprehensive reconstruction of their intricate 3D nanoporous architecture. This is precisely where computational methodologies emerge as a beacon of promise, unveiling their potential to elucidate experimentally observed phenomena and to decipher the intricate interplay between structure and properties.

In pursuit of this objective, the lecture will delve into the realm of computational design of nanoporous materials, with a special emphasis on fascinating aerogels [1]. To this end, the matter of the talk will encompass aerogels derived from inorganic sources, notably silica, as well as those rooted in organic origins, exemplified by phenolic-based variants. In the backdrop of the global quest for sustainable and environmentally responsible solutions, biobased materials, specifically those arising from biowaste, command growing attention. In this context, the meticulous modelling of such aerogels, particularly from polysaccharides such as cellulose and carrageenan, will be illustrated. Furthermore, the seamless integration of these models within a computational framework for the purpose of simulating composites will be discussed.

Finally, as we usher in the era of rapid materials development, the latter part of the discussion will turn its attention towards the harnessing of machine learning approaches. These approaches not only facilitate predictive modelling but also empower the reverse engineering of synthesis parameters, thereby paving the way for the advent of AI-driven, self-learning laboratories.