Application of CFD, statistical analysis and DoE for optimising H2 generation through ammonia borane catalytic hydrolysis: A novel approach for multiobjective optimisation
Journal
Chemical Engineering Journal
Date Issued
December 1, 2025
DOI
10.1016/j.cej.2025.170772
Abstract
A non-noble metal catalyst (Cu/NiO/TiO<inf>2</inf>) was used for the hydrolysis of ammonia borane (AB), studying the effect of temperature, AB concentration and substrate to catalyst (Sub:Cat) molar ratio. A zero-dimensional (0-D) model was developed using COMSOL Multiphysics for the validation of the experimental data, with a maximum error (17 %) observed at the highest studied concentration. Using the Isalos Data Analytics platform, a Box-Behnken Design (BBD) was implemented, studying complex interactions among the three studied parameters on the three critical response variables: reaction time, turnover frequency (TOF), and hydrogen (H<inf>2</inf>) yield. Main effects plots revealed nonlinear relationships between each factor and the mean of reaction time. Additional experiments were integrated into the statistical analysis, revealing how one parameter or interaction between the studied parameters influence the response values. Sub:Cat was found to be an influential term at all cases. A multiobjective optimisation analysis was conducted to identify optimal experimental conditions, showcasing a consistent trend favouring a temperature of 50 °C, Sub:Cat ratios around 1100–1150, and AB concentrations of 74–78 mM, generating H<inf>2</inf> yields above 74.4 % and reaction times below 30 min. Changing one-factor-at-a-time would have required 27 experiments, while with the use of BBD, only 13 experiments were needed to capture the main, quadratic, and interaction effects, reducing the experimental studies by almost 50 %, saving time and consumables.