A research team led by Associate Professor Sun Yifei from the School of Energy Science at Xiamen University, in collaboration with Professor Tu Xin from the University of Liverpool and Professor Shao Zongping from Curtin University, has published a groundbreaking study titled “Co-expression of multi-genes for polynary perovskite electrocatalysts for reversible solid oxide cells” in the prestigious international journal Nature Communications.
This study pioneers the application of a multi-gene co-expression strategy to systematically design the LnBaCo2O5+δ perovskite system, employing Bayesian optimization-based symbolic regression to decode material activity descriptors. The research reveals that the synergistic effects of configuration entropy (Sconfig), ionic radius (R), and electronegativity (χ) are crucial determinants of catalytic performance.
Solid oxide cells (SOCs) are highly efficient energy conversion devices that play a pivotal role in sustainable energy systems. However, traditional perovskite-type air electrodes face challenges such as limited activity, poor stability, and rapid electrochemical performance degradation, which hinder their widespread application.
The research team constructed a material library comprising over 50 perovskite variants, challenging the conventional single-descriptor material optimization approach and emphasizing the significance of multi-parameter synergistic effects in material design. Utilizing an innovative interpretable machine learning workflow, the team screened and validated three superior electrocatalysts from 177,100 unexplored compositions. Notably, (Pr0.05La0.4Nd0.2Sm0.1Y0.25)BaCo2O5+δ (PLNSY) demonstrated exceptional activity and stability in both solid oxide fuel cell (SOFC) and electrolyzer (SOEC) modes. Physical characterization and theoretical calculations revealed that PLNSY exhibits more uniform oxygen vacancy distribution, forming a 3D proton transport network with lower electrochemical energy barriers and higher stability.
This research establishes a quantitative framework linking entropy regulation engineering in complex oxides to catalytic functionality, transcending the limitations of single-descriptor material performance predictions and showcasing the immense potential of multi-gene co-expression strategies in material design. This methodology not only applies to the field of solid oxide cells but also offers novel insights for the development of other functional materials.
