According to the Royal Society of Chemistry, concrete is the single most widely used material in the world. However, the production of one of its main ingredients, cement, emits large amounts of carbon dioxide (CO₂) into the atmosphere.
This problem has inspired researchers from the Paul Scherrer Institute (PSI) in Switzerland to develop an AI-based model that speeds up the discovery of new cement recipes with better carbon footprints.
“It’s like having a digital cookbook for climate-friendly cement,” says first author of the study, mathematician Romana Boiger.
“Instead of testing thousands of variations in the lab, we can use our model to generate practical recipe suggestions within seconds”.
The research into this model was carried out as a part of the Swiss Centre of Excellence on Net Zero Emissions (SCENE) Project.
“Our method allows us to significantly accelerate the development cycle by selecting promising candidates for further experimental investigation,” says Nikolaos Prasianakis, who initiated the study.
Around 8% of the world’s CO₂ emissions are produced by the cement industry. In comparison, the aviation industry is responsible for 2.5% of the world’s carbon emissions.
The current cement production releases CO₂ emissions in two main ways.
Firstly, to make cement, raw materials like limestones are heated up in a giant oven known as a kiln. In order to transform the limestone into small pieces called clinker, the kilns are heated to 1,400°C. Heating the kilns up to such an intense level of heat requires both electricity and energy-intense combustion reactions which emit large amounts of the greenhouse gas.
But most of the CO₂ emissions produced in the production of cement actually come from the raw materials themselves. As limestone is transformed into clinker, the CO₂ that was chemically bound in the limestone is released.
This has led the team at PSI’s Centre for Nuclear Engineering and Sciences to try and modify the cement recipe’s molecules as a strategy to reduce emissions.
“With the help of the open-source thermodynamic modelling software GEMS, developed at PSI, we calculated – for various cement formulations – which minerals form during hardening and which geochemical processes take place,” says Nikolaos Prasianakis, who initiated the study.
The AI model created by the team is also able to calculate the total CO₂ emissions for different cement materials recipes in milliseconds.
“That is around a thousand times faster than with traditional modelling,” says Boiger.
“You could say we’re doing geology in fast motion,” says John Provis, head of the Cement Systems Research Group at PSI and co-author of the study.
“What we need is the right combination of materials that are available in large quantities and from which high-quality, reliable cement can be produced.”
While the researchers have discovered many promising candidates for new cement recipes, the materials now need to be tested in a laboratory first before they can be used in the construction industry.
“This is just the beginning,” says Nikolaos Prasianakis.
“The time savings offered by such a general workflow are enormous – making it a very promising approach for all sorts of material and system designs.”
The research is published in Materials and Structures.