Adsorptive Removal of Surfactants from Laundry Wastewater Using ML Data Analysis

Abstract

The release of surfactant-loaded laundry wastewater without proper treatment is a major environmental concern as it causes foam and reduces oxygen level in
aquatic bodies. Conventional methods are often not only expensive but also non-environmentally friendly. This paper explores a new bio-composite material,
Graphene Oxide modified Coconut Shell Activated Carbon (GO-CSAC), for the purpose of adsorbing surfactants effectively. GO-CSAC has outstanding structural
features and displays a very large surface area of 1120 m²/g along with a pore size of 1.75 nm. It uses the natural characteristics of activated carbon along with the
additional functional groups of graphene oxide to maximize adsorption sites. Various batch experiments were performed to check performance at different
conditions. Results show that the GO-CSAC is capable of adsorption much more than the unmodified carbon. Under ideal conditions which are the amount of
adsorbent = 1 g/L, time of contact = 6 hours, and pH value = 8, the material managed to remove 95% of surfactants. The Langmuir isotherm (R^2 = 0.995) and
pseudo-second-order model (R^2 = 0.965) fit to the adsorption data shows a close correlation between theoretical and experimental results. To make the operation
more efficient, a Machine Learning (ML) model was designed based on the experimental dataset to forecast the performance. This model gave very low prediction
error during validation thus drastically decreasing experimental trial-and-error. Making use of this environmentally friendly bio-composite produced from waste
and best-quality predictive modeling together is a very effective, intelligent solution for treating household wastewater and achieving pollution mitigation in a
sustainable way.