4D Early Age Cement Hydration Analysis by Ptychographic X-ray Computed Tomography and Machine Learning Segmentation

Abstract

Cement manufacturing is responsible for ~7% of the anthropogenic CO2 emissions and hence, decreasing the CO2 footprint, in a sustainable, safe and cost-effective way, is a top priority. To fully understand the binder main properties and to decrease their CO2 footprints, a sound description of their spatially resolved mineralogy is necessary. Developing this knowledge is very challenging as about half of the volume of hydrated cement is a nanocrystalline component, calcium silicate hydrate (C-S-H) gel. Furthermore, other poorly crystalline phases (e.g. iron siliceous hydrogarnet or silica oxide) coexist. Here, we have used ptychographic X-ray computed tomography (PXCT) for understanding the first days of cement hydration with the final goal to improve the mechanical strength performances of low-CO2 cements.