Transforming MOF Design Paradigms: Data-Driven Insights into MIL-100(Fe) Synthesis for Application-Specific Optimization of Dye Removal Performance

Traditional MOF design often maximizes generic metrics such as BET surface area and crystallinity, assuming they universally predict performance. We present a machine-learning-guided optimization framework for MIL-100(Fe) from experimentally synthesized samples. All materials were synthesized via an acid-free, water-based hydrothermal route. We trained small-data ML models to link synthesis parameters, including temperature, time, metal-to-ligand ratio, and ion concentration, to key properties comprising surface area, total pore volume, average crystallite size, crystallinity, yield, and methylene blue (MB) removal. SHAP analysis showed that time and metal-to-ligand ratio dominated dye removal, whereas surface area was more sensitive to temperature and time. The most accurate model, gaussian process regression was coupled with a genetic algorithm (GA) to optimize synthesis for property-specific targets. Through optimization, we achieved a 5.37% rise in BET surface area (1748\ to 1841.9 m²/g) and a 12.1% gain in MB removal (88.6 to 98.3%) over the dataset’s best results. The MB removal optimized sample with a surface area of 1274.3 m²/g and 17.4% crystallinity, delivers the highest MB uptake of 98.3%. While, optimized sample for surface area, reaches 1841.9 m²/g (about 44% higher) with 34.9% crystallinity but achieves only 85.1% removal, about 13% lower than optimized MB sample; indicating that even substantial increases in surface area do not govern adsorption performance. A t-SNE analysis further revealed distinct clusters in property space, with optimized samples occupying separate neighborhoods, elucidating that MOF synthesis should be tailored to the target application rather than a single metric such as surface area.