Experiments and nanoscale simulations of geopolymers: porosity and molecular structure

Cement production is responsible for approximately 9.5% of global anthropogenic CO2 emissions and as a consequence the increasing attention to sustainability issues has promoted research in alternative materials to cement paste, such as geopolymers. Current studies on geopolymers mainly focus on mechanical strength but research on their long-term mechanical performance and durability is still limited. Reducing the uncertainties around geopolymer durability requires a better understanding of the microstructural features and microscopic mechanisms that govern their chemo-mechanical behaviour. This can be pursued using simulations based on High Performance Computing (HPC), as previously experienced in the field of traditional cement. The aim of this study is to combine results from experimental analysis with HPC-based theoretical understanding of the fundamental mechanisms of material\u2019s reactivity and ageing. The molecular structure of N-A-S-H (sodium aluminosilicate hydrate) geopolymer is considered as a model of metakaolin geopolymers. First results on molecular simulations based on a simplified crystalline version of N-A-S-H with hydrosodalite structure and Al/Si = 1 are discussed. The simulations indicate a path to develop a more fundamental understanding of Na-Si and Na-OH concentration effect on the porous structure and on the skeletal density