Machine-learning-assisted catalytic performance predictions of binary alloy catalysts for glucose hydrogenation

Sorbitol production involves glucose hydrogenation, which requires co-adsorption of glucose and hydrogen on the catalyst surface for high catalytic performance. Binary alloy catalysts can modulate substrate adsorption to achieve better catalytic performance than Raney Ni. Herein, we established a pioneering DFT/ML approach to investigate the adsorption energies of glucose (ΔE GCHO ) and H atoms (ΔE H ) on 1155 binary alloy catalysts. The Light Gradient Boosting Machine (LGBM) algorithm proved the most effective ML model, predicting ΔE GCHO and ΔE H with R² values of 0.785 and 0.636, respectively. Microkinetic simulation demonstrated a correlation between catalytic activity and adsorption energy, revealing high-performance catalyst screening criteria as ΔE GCHO = −1.45 to −0.65 eV and ΔE H = −0.55–0.00 eV. Nine possible binary alloy catalysts with high predicted activity were identified, with Pd 3 Mg performing best. The present study highlights the potential of the DFT/ML-assisted approach in the development of efficient glucose hydrogenation binary alloy catalysts. • A DFT-based machine learning approach was established to investigate ΔE ads of glucose and H atom on bimetallic catalysts. • Microkinetic simulation was used to identify screening criteria and 9 high-performance bimetallic catalysts were found. • Electronic analysis showed a positive correlation between Pd’s D-band center and glucose's ΔE ads on Pd-based catalysts.