Not all biochar is equal: Feedstock-specific trade-offs for agricultural use under PFAS and trace contaminant realities

Research output: Contribution to journalArticlepeer-review

Abstract

Biochar use in agriculture and as a contaminant sorbent has moved unevenly from promise to practice. Benefits and risks depend on feedstock, production context and exposure. Using per- and polyfluoroalkyl substances (PFAS) as a model contaminant, we propose a feedstock-specific framework that distinguishes biochar derived from lignocellulosic crop and forestry residues and biochar from biosolids (sewage sludge). The first group addresses soil acidity and physical constraints and can be engineered for contaminant control, but routine testing for polycyclic aromatic hydrocarbons (PAHs) is essential, and short-chain PFAS should not be assumed to be effectively retained without verification. The second group recycles phosphorus and immobilises some metals, yet it concentrates heavy metals and, for food-producing land, requires verified PFAS destruction with effective off-gas treatment. Priorities include incorporating PFAS into certification where exposure is plausible; conducting field trials to monitor leachate and edible tissues in PFAS-impacted settings; specifying end-of-life pathways for sorbent applications; and guidance that is evidence-based and adaptable to regional soils, crops and exposure scenarios. A feedstock-specific approach can shift the debate from generic advocacy to credible, safe and scalable adoption aligned with standards, policy and practice.

Original languageEnglish (US)
Article number181483
JournalScience of the Total Environment
Volume1016
DOIs
StatePublished - Feb 15 2026

ASJC Scopus subject areas

  • Environmental Engineering
  • Environmental Chemistry
  • Waste Management and Disposal
  • Pollution

Fingerprint

Dive into the research topics of 'Not all biochar is equal: Feedstock-specific trade-offs for agricultural use under PFAS and trace contaminant realities'. Together they form a unique fingerprint.

Cite this