AI tech could boost large-scale green ammonia production, study finds
The study, conducted at the University of New South Wales in Australia and published in the peer-reviewed journal Small, used a machine learning platform to sift through more than 8,000 potential metal combinations for use as catalysts. The AI model shortlisted 28 high-potential candidates, accelerating a process that typically takes years or even decades.
The most effective material was found to be a liquid-state, high-entropy metal alloy composed of iron, bismuth, nickel, tin, and zinc. This unusual mix forms a stable and highly active catalyst that functions at room temperature.
In lab tests, the catalyst achieved nearly 100% Faradaic efficiency – meaning virtually all the electrical energy was used productively – and boosted ammonia production rates sevenfold compared with previous green methods. The process runs at just 25°C, requiring less than 10% of the energy used in the traditional Haber-Bosch process, which operates at over 400°C and contributes around 2% of global greenhouse gas emissions.
… to continue reading this article and more, please login, register for free, or consider subscribing to gasworld
Register today
You’ve reached your weekly limit to access free articles!
Want to keep reading?
Please register for free and create a profile to gain access to this full article and gasworld’s daily news.
For access to more content including our monthly digital magazines, subscriber-only features or columns and all our other gasworld archives, please consider subscribing.
Alternatively, you can continue reading more articles as a guest on Friday, 20th June at 10:29PM
