MIT researchers have developed a method to predict how well materials can move protons in clean energy devices and other advanced technologies.
Many advanced energy technologies, including fuel cells, electrolyzers, and some low-power electronics, rely on protons to carry electric charge, and their performance depends largely on how easily protons can move. Metal oxides can conduct protons at high temperatures above 400°C, but identifying materials that perform efficiently at lower temperatures has remained a significant challenge.
MIT researchers have developed a model that predicts proton movement across a wide range of metal oxides, identifying the key features that facilitate proton transfer and, for the first time, showing how the flexibility of oxide ions enhances this process, providing insights that could guide the design of more efficient energy materials using protons, which are lighter, smaller, and more abundant than common alternatives like lithium ions.
Researchers hypothesized that the flexibility of the oxide ion sublattices plays a key role in proton conduction. To test this, they developed a metric called “O…O fluctuation,” which measures changes in spacing between oxygen ions caused by phonons at finite temperature. They also compiled a dataset of other structural and chemical features affecting proton mobility and quantified their impact on proton conduction.
A model trained on these features identified the two most important factors for proton transfer: hydrogen bond length and oxygen sublattice flexibility. Materials with shorter hydrogen bonds transported protons more efficiently, confirming previous studies. The O…O fluctuation metric was the second most important factor, showing that more flexible oxygen ion chains improve proton conduction.
This model can estimate proton conduction across a broad range of materials and could help identify promising candidates for energy and computing applications. It can also guide generative AI models to design new materials optimized for proton transfer, potentially enabling a new class of highly efficient clean energy technologies. Future work will focus on understanding which compositions and structures create flexible, percolated oxide ion sublattices to further improve performance.
