A study published in Environmental Science and Ecotechnology reports a self-powered platform that integrates electrically assisted forward osmosis (eFO) with a microbial desalination cell (MDC) to recover water and fertilizer from livestock wastewater without external electricity. The system uses microorganisms to generate electricity from organic matter, which then drives ion migration and struvite precipitation, producing a slow-release fertilizer while improving water flux and desalination.
Conventional wastewater treatment has largely focused on pollutant removal, but growing water scarcity, energy constraints, and fertilizer demand are pushing the field toward resource recovery. Forward osmosis (FO) can draw water across a membrane with low hydraulic pressure, while bioelectrochemical systems such as microbial desalination cells can convert organic matter into electricity and help move salts. Yet these tools are often run separately: FO systems still face concentration polarization and reverse salt flux, and electrically assisted FO usually needs continuous external power supply. The new study addresses these challenges by combining the two technologies in a closed-loop design.
Researchers from Temple University and New Jersey Institute of Technology published the study (DOI: 10.1016/j.ese.2026.100730) in Environmental Science and Ecotechnology, accepted on July 4, 2026. The integrated system uses bioelectricity generated during organic matter oxidation to drive ion migration, recover struvite fertilizer, improve water flux, and enhance desalination from synthetic livestock wastewater.
In the eFO module, an osmotic gradient pulls water from the wastewater side toward a magnesium sulfate draw solution. When a mild electric field is applied, magnesium ions migrate back toward the wastewater side, where they react with ammonium and phosphate to precipitate as struvite (magnesium ammonium phosphate hexahydrate), a slow-release fertilizer. In the MDC, electroactive microorganisms oxidize organic matter, generate electrons, and support desalination. The researchers harvested this microbial electricity, stored it in a 400-farad supercapacitor, regulated the voltage, and fed it back to the eFO unit. At bench scale, the MDC generated more than 7.0 milliwatts, while the eFO module consumed less than 1.0 milliwatt. Compared with the control, water flux rose by 57%, struvite recovery increased from 0.25 to 0.71 grams at 1.8 volts, and total desalination efficiency improved by 45%. At higher voltage, struvite recovery reached 1.03 grams at 3.8 volts.
To guide operation, the team also developed a hybrid model that combined mechanistic transport equations with a support vector machine (SVM), enabling prediction of struvite recovery, chemical oxygen demand (COD), conductivity, and power output across different operating conditions. The authors said the study shows how wastewater treatment can be redesigned as a connected resource-recovery loop rather than a set of separate unit operations. They noted that the important step lay not only in coupling a membrane process with a bioelectrochemical process, but also in allowing the electricity generated by microorganisms to directly control ion movement and fertilizer formation. In their view, this internal feedback makes the approach more practical for nutrient-rich streams such as livestock wastewater, where water recovery, salinity control, and phosphorus recovery can all create value.
The results point to applications in decentralized wastewater treatment, agricultural waste management, and future resource-recovery facilities. The paper also makes clear that scale-up will require engineering work: the MDC produced enough power for the eFO module, but hydraulic retention times, module sizing, struvite harvesting, membrane scaling, and electrode durability still need optimization. A techno-economic assessment (TEA) estimated a bench-scale net treatment cost of $10.2 per cubic meter, falling to $3.3 per cubic meter in an engineering scale-up scenario. The study was partially supported by the U.S. Bureau of Reclamation (Award#: 13761566 and R22AC00433) and the NSF/BSF project (Award#: 2215387). Publication of this article was funded in part by the Temple University Libraries Open Access Publishing Fund.
