Effective Pretreatment of Marine Macroalgal Biomass (Ulva reticulata) for Enhanced Biohydrogen Production: An Artificial Intelligence-Assisted Evaluation

The limited accessibility of fermentable substrates in marine macroalgal biomass constrains efficient biohydrogen production. This study evaluated thermal, disperser and biosurfactant pretreatments individually and in combination to enhance Ulva reticulata biomass solubilization and biohydrogen production. Thermal (50–90 °C), disperser (4000–14,000 rpm) and biosurfactant (0.002–0.09 g/g total solids) pretreatments were investigated. Thermal pretreatment (80 °C, 45 min) achieved 2640 mg/L soluble chemical oxygen demand (sCOD) release at a specific energy requirement of 44,064 kJ/kg total solids (TS), whereas disperser pretreatment (8000 rpm, 30 min) released 2720 mg/L sCOD at 31,347 kJ/kg TS. Integrated thermal–disperser–biosurfactant pretreatment using 0.01 g/g TS biosurfactant achieved the highest sCOD release (3015 mg/L) and solubilization (26%) at 20,898 kJ/kg TS. Dark fermentation using camel dung as inoculum produced a maximum yield of 83 mL H2/g COD, compared with 50 and 58 mL H2/g COD for the thermal and thermal–disperser pretreatments, respectively. Firmicutes was the most abundant phylum, with Romboutsia, Clostridium sensu stricto 1, Turicibacter, and Paeniclostridium as the major genera. An artificial neural network accurately reproduced cumulative biohydrogen trends under leakage-free validation and ranked pretreatments according to experimental results. Overall, integrated pretreatment enhanced biomass solubilization and biohydrogen production while reducing the energy requirement.