Persistent storage of captured CO2 is essential to reduce the amount of CO2 in the atmosphere. Basalts have been suggested as safe and efficient for carbon storage due to their widespread occurrence (~10 % of the continents) and their reactivity with carbonated water. Cations in the basalt react with CO2 to form carbonate minerals, trapping CO2 in a solid phase.
One of the important mechanisms for creating reactive basalt surfaces, and to keep fluid pathways open during injection, is fracturing. Understanding the fracture properties of basalts is therefore important to assessing how CO2 can be efficiently stored in basalts. In this proposal, the project seeks the computing time to simulate the fracture process in basaltic glass on the molecular scale. T
o achieve the necessary accuracy for fracture simulations, and to reach the length- and timescale needed to capture relevant fracture properties, the project will use highly scalable state-of-the-art neural network based molecular dynamics simulations trained on DFT calculations.
Using these simulations, it will determine how phenomena like crack oscillations and branching affect reactivity and connectivity. As a secondary goal, the project will assess whether a theory for sub-critical fast crack propagation for scotch tape and PMMA also works for glasses.
University of Oslo, Norway.