Skip to main navigation Skip to search Skip to main content

Electrocoagulation as a high-efficiency strategy for microplastic removal from water: Process optimization and predictive modeling

  • Universidad Tecnológica del Perú
  • Universidad San Ignacio de Loyola
  • Departamento de Quimica Aplicada a Sistemas de Produccion Universidad de Cuenca

Research output: Contribution to journalArticlepeer-review

Abstract

This study evaluated and optimized the removal of polyester microplastics (MPs) from water by electrocoagulation (EC), while developing a predictive model and estimating operating cost. Synthetic effluent containing polyester MPs (51–150 µm) was treated in a batch EC system using iron electrodes, direct current, magnetic stirring, and NaCl as a supporting electrolyte. A full factorial 25 design (32 runs) was first applied to assess the effects of pH, voltage, operation time, NaCl concentration, and initial MPs concentration. Subsequently, a Box–Behnken design (15 runs) was used to optimize the most influential variables. Screening experiments showed high removal efficiencies ranging from 80% to 100%, identifying MPs concentration, NaCl concentration, and operation time as the dominant factors. The regression model showed strong predictive performance, with R2 = 0.9852 and adjusted R2 = 0.9655. Under optimized conditions of 200 mg L−1 MPs, 363.64 mg L−1 NaCl, and 18.38 min, complete MPs removal was achieved, with an estimated operating cost of US$0.57 m−3. These results indicate that EC is an efficient and low-cost strategy for polyester MPs removal, although validation in real effluents remains necessary.

Original languageEnglish
Article number101812
JournalDesalination and Water Treatment
Volume326
DOIs
StatePublished - Jan 2026

Keywords

  • Electrocoagulation
  • Microplastics
  • Polyester
  • Response surface methodology
  • Wastewater

Fingerprint

Dive into the research topics of 'Electrocoagulation as a high-efficiency strategy for microplastic removal from water: Process optimization and predictive modeling'. Together they form a unique fingerprint.

Cite this