“`html

Modern civilization relies heavily on concrete: Humans consume more concrete each year than any other substance apart from water. However, cement, the fundamental element of concrete, is responsible for up to 10% of global carbon dioxide emissions.

To tackle this issue, scholars at the University of Washington and Microsoft have created a novel type of low-carbon concrete by blending dried, powdered seaweed with cement. This seaweed-enriched cement has a 21% reduced global warming potential while maintaining its strength. With the aid of machine learning models, the team developed this innovative formula in a fraction of the time typically required for such research.

The team released its results on July 8 in Matter.

“Cement is ubiquitous — it underpins modern infrastructure — yet it carries a significant climate burden,” stated senior author Eleftheria Roumeli, a UW assistant professor in materials science and engineering. “What makes this project thrilling is that we demonstrate how a plentiful, photosynthetic resource like green seaweed can be integrated into cement to reduce emissions, without requiring expensive processing or compromising efficacy.”

Producing one kilogram of cement releases nearly one kilogram of CO2. The majority of these emissions originate from the fossil fuels utilized to heat raw materials and from a chemical reaction called calcination that occurs during manufacturing. In contrast, seaweed acts as a carbon sink: it absorbs carbon from the atmosphere and retains it as it grows. Remarkably, it can directly substitute a portion of the cement in concrete, resulting in a significantly smaller carbon footprint.

Determining the optimal blend of components might have taken five years of experimentation, Roumeli estimated, since each concrete sample requires about a month to completely cure before its properties can be accurately assessed.

To expedite the process, the team constructed a tailored machine learning model and trained it using an initial set of 24 cement formulations. They then employed the model to predict ideal mixtures for laboratory testing. By inputting the test results back into the model, they collaborated with the model and quickly moved through formulations. The result was an ideal mixture of seaweed-infused cement with a lowered carbon footprint that passed compressive strength assessments, achieved in just 28 days.

“Machine learning was crucial in allowing us to significantly shorten the process — especially vital here, because we’re integrating a completely new substance into cement,” Roumeli noted.

Going forward, the team intends to deepen their comprehension of how the composition and structure of seaweed influences cement performance. The broader aim is to extend this work to various types of algae (or even food waste) so that producers can develop localized, sustainable cement alternatives globally — and utilize machine learning to optimize them swiftly.

“By merging natural materials like algae with contemporary data tools, we can localize production, diminish emissions, and accelerate the shift towards greener infrastructure,” Roumeli added. “It represents an exciting advancement toward a new generation of sustainable building materials.”

Additional co-authors on this study include Meng-Yen Lin, a UW doctoral student in materials science and engineering; Paul Grandgeorge, a former UW postdoctoral researcher in the materials science and engineering department who is now an R&D engineer at the iPrint Institute; and Kristen Severson, a principal researcher at Microsoft Research.

This study was funded by Microsoft Research.

For further details, contact Roumeli at [email protected].

“`