Modelling of an adaptive neuro-fuzzy inference system and artificial neural network for the removal of cobalt (II) from aqueous solutions using green and biodegradable adsorbents: Column breakthrough studies
- 1Corresponding author.
- 2Centre for Integrated Water Resource Management Department of Environmental Science, School of Ecological and Human Sustainability, College of Agriculture and Environmental Sciences, University of South Africa, Christiaan de Wet Road, FL, 1710, South Africa
Abstract
This study investigates the continuous removal of cobalt (II) from water using a novel, environmentally friendly, and renewable adsorbent based on modified cellulose nanocrystal (CNC) composites. Fixed-bed column experiments were conducted at different pH values (4–8), influent cobalt (II) concentrations (10–20 mg/L), flow rates (3–7 mL/min), and bed heights (3–9 cm) to evaluate breakthrough behavior and column performance. The novelty of this study lies in integrating continuous fixed-bed adsorption experiments, conventional breakthrough modelling, and artificial intelligence-based prediction using artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) models within a unified framework. Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), and thermogravimetric analysis (TGA) confirmed successful adsorbent modification and favorable physicochemical characteristics. Column performance decreased with decreasing bed height and increasing flow rate and influent concentration, primarily because of reduced contact time and faster saturation of available adsorption sites. Among the conventional breakthrough models investigated, the Yan model provided the best representation of the nonlinear adsorption behavior. ANN and ANFIS models were subsequently developed to predict cobalt (II) removal under dynamic operating conditions and evaluated using R², RMSE, MSE, and MPSD. The ANN model demonstrated excellent predictive performance, achieving R² = 0.999 and strong agreement between experimental and predicted results. Integrating renewable CNC-based adsorbents with conventional modelling and ANN/ANFIS prediction provides a promising framework for optimizing continuous cobalt (II) removal and generating complementary information to support engineering scale-up and advanced water-treatment applications.
Related
- Artificial neural network-assisted modelling of methylene blue dye adsorption onto ZIF-8 functionalized fishtail palm seed biomass: Isotherm and performance analysis
- Preparation of Customizable Ligand-Functionalized Chitin Biopolymer Adsorbents Utilizing the Azide–Alkyne “Click” Reaction for the Adsorption of Aqueous Cu(II)
- Upcycling Used Tyre Rubber into Carbon Adsorbents: Effect of Physical Activation on Porous Structure and Comparative Evaluation for p-Chlorophenol Removal
- Ziziphus Seed-Based Adsorbents: A Review Study on Preparation, Adsorption Mechanisms, and Environmental Applications
- The plastisphere: from microbial pollution to biodegradable solution
BibTeX
@article{claude2026modelling,
title = {{Modelling of an adaptive neuro-fuzzy inference system and artificial neural network for the removal of cobalt (II) from aqueous solutions using green and biodegradable adsorbents: Column breakthrough studies}},
author = {Banza Jean Claude and Vhahangwele Masindi and Alusani Manyatshe},
journal = {Desalination and Water Treatment},
year = {2026},
month = oct,
doi = {10.1016/j.dwt.2026.102016},
url = {https://doi.org/10.1016/j.dwt.2026.102016},
note = {CC BY},
}
RIS
TY - JOUR
TI - Modelling of an adaptive neuro-fuzzy inference system and artificial neural network for the removal of cobalt (II) from aqueous solutions using green and biodegradable adsorbents: Column breakthrough studies
AU - Banza Jean Claude
AU - Vhahangwele Masindi
AU - Alusani Manyatshe
PY - 2026
DA - 2026/10/01
JO - Desalination and Water Treatment
DO - 10.1016/j.dwt.2026.102016
UR - https://doi.org/10.1016/j.dwt.2026.102016
AB - This study investigates the continuous removal of cobalt (II) from water using a novel, environmentally friendly, and renewable adsorbent based on modified cellulose nanocrystal (CNC) composites. Fixed-bed column experiments were conducted at different pH values (4–8), influent cobalt (II) concentrations (10–20 mg/L), flow rates (3–7 mL/min), and bed heights (3–9 cm) to evaluate breakthrough behavior and column performance. The novelty of this study lies in integrating continuous fixed-bed adsorption experiments, conventional breakthrough modelling, and artificial intelligence-based prediction using artificial neural network (ANN) and adaptive neuro-fuzzy inference system (ANFIS) models within a unified framework. Fourier transform infrared spectroscopy (FTIR), scanning electron microscopy (SEM), and thermogravimetric analysis (TGA) confirmed successful adsorbent modification and favorable physicochemical characteristics. Column performance decreased with decreasing bed height and increasing flow rate and influent concentration, primarily because of reduced contact time and faster saturation of available adsorption sites. Among the conventional breakthrough models investigated, the Yan model provided the best representation of the nonlinear adsorption behavior. ANN and ANFIS models were subsequently developed to predict cobalt (II) removal under dynamic operating conditions and evaluated using R², RMSE, MSE, and MPSD. The ANN model demonstrated excellent predictive performance, achieving R² = 0.999 and strong agreement between experimental and predicted results. Integrating renewable CNC-based adsorbents with conventional modelling and ANN/ANFIS prediction provides a promising framework for optimizing continuous cobalt (II) removal and generating complementary information to support engineering scale-up and advanced water-treatment applications.
SN - 1944-3986
SN - 1944-3994
ER -