Algorithm finds best carbon dioxide 'sponges' from a trillion recipes
Researchers borrowed from Darwinian principles to identify the promising carbon capture metallic-organic frameworks (MOFs). Angus Bezzina reports.
Carbon dioxide “sponges” called metallic-organic frameworks, or MOFs, soak up the greenhouse gas before it's emitted. But with almost limitless combinations of ingredients, how do materials scientists hit on the best MOF recipe?
A Canadian team took inspiration from evolution to whittle more than a trillion recipes down to less than 200 of the most promising. Sean Collins and colleagues at the University of Ottawa tweaked an algorithm used in genetic studies to cross MOF “parent” structures and find the most effective “daughter” materials.
When the algorithm trawled through 1.65 trillion potential MOF molecular structures, it turned up more than 1,000 “exceptional” MOFs and of those, 141 that could be synthesised and tested.
The work, which was published in Science Advances, not only looks at the gas-capturing qualities of the materials, but also the amount of energy needed to release it again for storage – called “parasitic energy” – which materials scientist Matthew Hill at Australia’s Commonwealth Scientific and Industrial Research Organisation says is as important as absorbency.
But he is quick to caution that there are still a number of other factors that need to be examined in the lab before any of the MOFs identified in the study can be implemented in a modern industrial setting.
Renewable energy is on the rise, with solar power tipped to become the cheapest form of electricity by 2025. But many countries still heavily rely on burning coal and other fossil fuels – and this comes with a price.
Carbon dioxide, produced as coal burns, is usually released into the atmosphere where it contributes to global warming.
Finding ways to nab the gas between combustion and emission is a goal for many research groups – and MOFs are a promising area.
“Inside that [the MOF] it kind of looks like a building scaffold so that the corners are a metal atom or a group of metal atoms,” Hill, who works with MOFs, explains.
"Then the struts joining those corners to one another are organic molecules and it builds up into a big, long structure."
Carbon dioxide molecules lodge and are held in gaps in the crystal structure, and released when the material is, for instance, heated.
MOFs can be made from an almost limitless number of combinations of metal atoms and organic molecules and fine-tuned by tailoring molecular groups within them.
But given the sheer number of these possible marriages, finding the optimal MOF pairing can be difficult.
This is where the genetic algorithm comes into play.
Inspired by Darwinian evolution, such algorithms use a process akin to natural selection to find the “fittest” combinations.
In this case, Collins and his team started with 23 parent MOFs, which researchers have already made and tested.
These were “mated” and the daughter MOFs judged based on three properties: carbon dioxide intake, surface area (more surface area means more absorbency) and parasitic energy – that is, the energy needed to heat the MOF so it relinquishes its hold on the carbon dioxide molecules.
The average carbon dioxide uptake rate increased drastically – but a factor of almost four – and 1,035 derivatives had “exceptional uptake” of carbon dioxide.
Hill says the next steps should be to experimentally show the carbon dioxide uptake of the most promising structures, prove their stability in industrial environments and present a simple, safe and low-cost method for mass production.