Water hyacinth-derived adsorbent for removal of diclofenac sodium from aqueous solution: Modeling and optimization via RSM-ANN/ANFIS hybrid
Over-the-counter medications, mainly diclofenac sodium (DCFS), have emerged as a huge business in the pharmaceutical industry. Consequently, this has led to the widespread presence of DCFS in water bodies, posing adverse effects on the entire biota. Also, the rapid growth of invasive water hyacinth...
Saved in:
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Elsevier
2025-10-01
|
Series: | Next Materials |
Subjects: | |
Online Access: | http://www.sciencedirect.com/science/article/pii/S2949822825005155 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | Over-the-counter medications, mainly diclofenac sodium (DCFS), have emerged as a huge business in the pharmaceutical industry. Consequently, this has led to the widespread presence of DCFS in water bodies, posing adverse effects on the entire biota. Also, the rapid growth of invasive water hyacinth (WH) has continuously impeded aquatic physiological processes and even fishing activities. Thus, this study investigated the WH as the precursor of activated carbon (WH-AC) for the adsorption of DCFS from an aqueous solution. The characterization analysis confirmed the successful synthesis of facile WH-AC. The RSM optimum condition for 100 % efficiency was determined to be 1.191 g/L dosage, 21.064 mg/L initial concentration, and 107 min contact time. Under the same conditions, the experimental efficiency was determined to be 97.9 %, which validates the RSM quadratic model. In addition, the adsorption of DCFS onto the WH-AC was determined to follow D-R>Langmuir>Temkin>Freundlich isotherms. The D-R (R² = 0.9916) and E (2.5 ×10⁻⁶ kJ/mol) less than 10 kJ/mol imply that the adsorption mechanism is physical with Gaussian energy distribution over heterogeneous surfaces. Furthermore, a positive Temkin adsorption energy (BT = 14.5 J/mol) and negative ΔH° (-21841 Jmol−1) indicate that the adsorption is exothermic. In addition, the ANN model using trainLM with log-sigmoid showed the best adsorption efficiency prediction compared to ANFIS based on speed and statistic indicators. In conclusion, ANN can give a heuristic prediction necessary for making critical decisions about the adsorption process. Also, the utilization of WH-AC can guarantee a green solution for WH management and water purification. |
---|---|
ISSN: | 2949-8228 |