Data-driven framework for biochar engineering: Balancing heavy metal risk, carbon purity, and feedstock-pyrolysis trade-offs

Up-scaling biochar production requires managing integral trade-offs. While higher pyrolysis temperature increases carbon stability, it also concentrates ash and heavy metals via mass loss. Existing frameworks often overlook this enrichment effect, underestimating ecological risk. Without generalized boundaries defining how different feedstocks and operational conditions balance these competing factors, safe and optimized biochar design remains elusive. Here, a cross-study dataset comprising 167 publications (1119 samples) was synthesized to establish interpretable, data-driven boundaries for biochar design via slow pyrolysis, which prioritizes empirical ranges and consistent drivers over black-box point predictions. Four feedstock categories (agro, wood, manure, and sludge) were systematically compared. Pyrolysis temperature was identified as the primary operational parameter, while heating rate and residence time exerted secondary effects. Heavy metal risk was reassessed by integrating each metal's ecological risk factor with its concentration factor. An enrichment-corrected risk index was developed: at ≥ 600 °C, sludge-biochar retained about 22% of raw sludge's risk index, driven primarily by Cd and Cu immobilization. Besides, a carbon purity index (− 1 to 1) was proposed, revealing strong feedstock-dependent stratification: agro (0.58 ± 0.30), wood (0.85 ± 0.19), manure (− 0.04 ± 0.29), and sludge (− 0.48 ± 0.26). This framework provides transparent, mechanistically interpretable guidance for feedstock selection, process optimization, and risk management, offering an empirical basis for engineering decision-making and future industrial validation.