How many times have we held our breath while passing by an old diesel, knowing that those fumes do not promise anything good? Nitrogen oxides are among the worst enemies of our lungs, silent and invisible aggressors that the Euro 7 regulations finally want to confine to the enclosure of the โalmost eliminatedโ.
But how can this goal be achieved without declaring war on four wheels? The answer comes from a team of Chalmers University, who has come up with an innovative method: deploying artificial intelligence to design dramatically more efficient copper-zeolite catalysts. It's mathematics teaming up with chemistry to allow us to continue driving without guilt.
Euro 7, a question of standards and innovation
Euro 7 regulations loom on the horizon as a formidable challenge for automakers. This is not a small adjustment: it is a further, drastic reduction in the pollutants allowed in emissions. Old catalytic converters, however effective, may not be up to the task.
And this is where innovation comes into play: catalysts based on copper-enriched chabazite zeolite. These materials have demonstrated surprising effectiveness in the selective catalytic reduction of nitrogen oxides, using ammonia as the reducing agent. The process is fascinating: they promote the formation of nitrogen-nitrogen bonds in an oxygen-rich environment, while preventing unwanted reactions.
The complexity of these systems is such that only artificial intelligence can help us truly understand them. No wonder scientists have turned to you to solve this molecular puzzle.
Inside the microscopic world of catalysts
The magic of these catalysts lies in their extraordinarily dynamic nature. Zeolites are like tiny crystalline cages, where copper ions dance a molecular waltz with ammonia, forming mobile complexes that float in the channels of the material.
Computational investigations are important to understand how detailed structure and composition influence their performance.
Words of Henrik Groenbeck, professor at the Department of Physics of the Chalmers University of Technology. The mobility of these complexes is crucial: two complexes in the same zeolite cage are needed for the reaction to proceed. Like bringing together two dancers in a crowded room. AI has the answer.
The research team developed a machine-learning โforce field,โ a computational model that describes the forces between atoms, including long-range electrostatic interactions. This allowed them to study the diffusion of charged ammonia-copper-ammonia complexes, providing insights that were previously inaccessible.
The impact of simulations
I study, published Nature Communications., bears the signature of Henrik Groenbeck, Joachim Bjerregaard (PhD student at the Department of Physics) and Martin Votsmeier (industrial partner Umicore and Darmstadt University of Technology), within the framework of the CHASS project.
I like to think that while we are debating the restrictions of Euro 7, little copper atoms are already learning to better purify the air, guided by artificial intelligence. Maybe, for once, the fact that technology arrives before the regulations will be useful to us.
