Rosa damascena waste as biosorbent for co-existing pollutants removal: fixed-bed column study and ANN modeling

• Main Authors: Batool, Fatima, Kurniawan, Tonni Agustiono, Mohyuddin, Ayesha, Othman, Mohd Hafiz Dzarfan, Ali, Imran, Abdulkareem-Alsultan, G., Anouzla, Abdelkader, Goh, Hui Hwang, Zhang, Dongdong, Aziz, Faissal, Wayne Chew, Kit
• Format: Article
• Published: Elsevier 2024
• Online Access: http://psasir.upm.edu.my/id/eprint/112801/; https://www.sciencedirect.com/science/article/abs/pii/S0009250924003579?via%3Dihub

The adsorption behavior of Cd(II) and Pb(II) ions in a coexisting environment, alongside with both dyes (Reactive Red198 and Blue29) was systematically investigated by using a continuous fixed-bed column completely packed with Rosa damascena waste biosorbent (RDWB). Artificial neural network (ANN) model was also utilized to predict the performance of RDWB for various inputs. Their column performance was assessed by optimizing parameters such as bed depth, influent flow rate, and biosorbents’ concentration. It was found that an increasing bed depth considerably extended the operational lifespan and decrease in flow rate delays the column adsorption. Its adsorption capacities were 24.9, 24.6, 24.0 and 24.3 mg/g for Pb(II), Cd(II), Red198, and Blue29, respectively. The RDWB also had a higher adsorption capacity, as compared to previously used biosorbents such as chitosan and biochars due to its good thermal stability and high surface area of 421.46 m2/g. The adsorption of target pollutants took place through ion exchange and electrostatic interactions with negatively charged functional groups on the adsorbent's surface. The experimental data were fitted by various column adsorption models such as the Thomas, Yoon-Nelson, and Adams-Bohart. The findings showed that the Thomas model exhibited a strong correlation with the experimental data. In contrast, the Adams-Bohart model was applicable to the initial phase of the breakthrough curve (Ce/C0 ≤ 0.1). For industrial applications, a scale up model was also presented with the cost analysis of the biosorbent. The comparison of predicted values with experimental percentage (%) removal values of target pollutants by the RDWB indicated the excellent performance of the ANN model for this work.