For weeks, the whiteboard within the lab was crowded with scribbles, diagrams, and chemical formulas. A research team across the Olivetti Group and the MIT Concrete Sustainability Hub (CSHub) was working intensely on a key problem: How can we reduce the quantity of cement in concrete to avoid wasting on costs and emissions?
The query was actually not latest; materials like fly ash, a byproduct of coal production, and slag, a byproduct of steelmaking, have long been used to interchange among the cement in concrete mixes. Nonetheless, the demand for these products is outpacing supply as industry looks to cut back its climate impacts by expanding their use, making the seek for alternatives urgent. The challenge that the team discovered wasn’t a scarcity of candidates; the issue was that there have been too many to sort through.
On May 17, the team, led by postdoc Soroush Mahjoubi, published an open-access paper in Nature’s outlining their solution. “We realized that AI was the important thing to moving forward,” notes Mahjoubi. “There’s a lot data on the market on potential materials — a whole bunch of 1000’s of pages of scientific literature. Sorting through them would have taken many lifetimes of labor, by which period more materials would have been discovered!”
With large language models, just like the chatbots a lot of us use each day, the team built a machine-learning framework that evaluates and sorts candidate materials based on their physical and chemical properties.
“First, there may be hydraulic reactivity. The explanation that concrete is robust is that cement — the ‘glue’ that holds it together — hardens when exposed to water. So, if we replace this glue, we’d like to make certain the substitute reacts similarly,” explains Mahjoubi. “Second, there may be pozzolanicity. That is when a cloth reacts with calcium hydroxide, a byproduct created when cement meets water, to make the concrete harder and stronger over time. We want to balance the hydraulic and pozzolanic materials in the combination so the concrete performs at its best.”
Analyzing scientific literature and over 1 million rock samples, the team used the framework to sort candidate materials into 19 types, starting from biomass to mining byproducts to demolished construction materials. Mahjoubi and his team found that suitable materials were available globally — and, more impressively, many may very well be incorporated into concrete mixes just by grinding them. This implies it’s possible to extract emissions and price savings without much additional processing.
“A number of the most interesting materials that might replace a portion of cement are ceramics,” notes Mahjoubi. “Old tiles, bricks, pottery — all these materials can have high reactivity. That’s something we’ve observed in ancient Roman concrete, where ceramics were added to assist waterproof structures. I’ve had many interesting conversations on this with Professor Admir Masic, who leads a whole lot of the traditional concrete studies here at MIT.”
The potential of on a regular basis materials like ceramics and industrial materials like mine tailings is an example of how materials like concrete can assist enable a circular economy. By identifying and repurposing materials that may otherwise find yourself in landfills, researchers and industry can assist to provide these materials a second life as a part of our buildings and infrastructure.
Looking ahead, the research team is planning to upgrade the framework to be able to assessing much more materials, while experimentally validating a few of the perfect candidates. “AI tools have gotten this research far in a short while, and we’re excited to see how the most recent developments in large language models enable the following steps,” says Professor Elsa Olivetti, senior creator on the work and member of the MIT Department of Materials Science and Engineering. She serves as an MIT Climate Project mission director, a CSHub principal investigator, and the leader of the Olivetti Group.
“Concrete is the backbone of the built environment,” says Randolph Kirchain, co-author and CSHub director. “By applying data science and AI tools to material design, we hope to support industry efforts to construct more sustainably, without compromising on strength, safety, or durability.
Along with Mahjoubi, Olivetti, and Kirchain, co-authors on the work include MIT postdoc Vineeth Venugopal, Ipek Bensu Manav SM ’21, PhD ’24; and CSHub Deputy Director Hessam AzariJafari.