Catalytic peroxidation of winery wastewater contaminants using activated carbon-supported magnetite nanoparticles

In this study, co-precipitation and hydrothermal synthesis methods were employed to prepare two types of magnetite nanoparticles (MNPs). An activated carbon derived from olive stones (OSAC) was used as MNPs-support. The physicochemical properties of the resulting OSAC-MNPs catalysts were determined using various characterization techniques. A more homogeneous distribution of larger and well-defined MNPs was obtained by hydrothermal synthesis, resulting in a more extensive blockage of the microporous structure of the support. Both catalysts exhibited paramagnetic behavior, but with relatively low magnetization due to the small size and low crystallinity of the Fe3O4 nanoparticles obtained. The catalytic peroxidation performance of OSAC-MNPs was studied using tyrosol (TY) as a model compound, but also real samples of winery wastewater (WW) were used in the optimized operational conditions, including pH, temperature, and doses of catalyst and hydrogen peroxide. The MNPs-based catalyst prepared by co-precipitation performed better in the range of experimental conditions tested because of the higher surface concentration of active phase that is easily accessible to the pollutant and was less prone to deactivation during successive reaction runs. The combination of active materials and optimized process enables to remove up to 92 % TPh, 35 % COD, and 26 % TOC, while the high stability of the samples associated to a low Fe-leaching (0.07 mg L−1) permits the reuse of materials in consecutive cycles. •An agricultural by-product – olive stone – is integrated in winery wastewater treatment•Co-precipitation and hydrothermal methods are employed for MNP-catalysts synthesis•Tyrosol oxidation is correlated with the chemical & morphological properties of the materials•Fe surface distribution and size is key for MNP-catalysts' peroxidation efficiency•Consistently low iron leaching obtained in real wastewater treatment experiments