Researchers in Singapore have taken a deep dive into spinel oxides – a class of materials known to act as a catalyst in the production of hydrogen through water electrolysis. Better understanding of how the materials work enabled the scientists to develop a machine learning model to predict their efficiency.
While many analysts believe hydrogen has a vital role to play in the clean energy transition – and commercial applications are already taking shape – the electrolyzers required to produce it from renewable power remain expensive to operate, thanks in part to a need for precious metals to catalyze the chemical reactions that split water into hydrogen and oxygen.
With researchers on the hunt for cheaper alternatives, promise has been shown by spinel oxides – materials with a particular crystalline structure. However, understanding about how they work as an electrolysis catalyst is limited.
the researchers were able to develop a machine learning model to predict the effectiveness of spinel oxide catalysts. The model highlighted one – based on manganese and aluminum – from a list of more than 300 materials and the group confirmed its superior performance in experiments.
NTU said the paper sets mechanical principles for spinel oxides which could extend over a range of applications.
Read original full article
