Extensive sorption of Amoxicillin by highly efficient carbon-based adsorbent from palm kernel: Artificial neural network modeling

Khadijeh Jafari; Mohsen Heidari; Ali Fatehizadeh; Kavoos Dindarloo; Vali Alipour; Omid Rahmanian

doi:10.1016/j.heliyon.2023.e18635

Extensive sorption of Amoxicillin by highly efficient carbon-based adsorbent from palm kernel: Artificial neural network modeling

Heliyon. 2023 Jul 24;9(8):e18635. doi: 10.1016/j.heliyon.2023.e18635. eCollection 2023 Aug.

Authors

Khadijeh Jafari¹, Mohsen Heidari², Ali Fatehizadeh^{1

3}, Kavoos Dindarloo⁴, Vali Alipour⁴, Omid Rahmanian⁴

Affiliations

¹ Environment Research Center, Research Institute for Primordial Prevention of Non-Communicable Disease, Isfahan University of Medical Sciences, Isfahan, Iran.
² Department of Environmental Health Engineering, Faculty of Medical Sciences, Tarbiat Modares University, Tehran, Iran.
³ Department of Environmental Health Engineering, School of Health, Isfahan University of Medical Sciences, Isfahan, Iran.
⁴ Department of Environmental Health Engineering, Faculty of Health, Hormozgan University of Medical Sciences, Bandar Abbas, Iran.

Abstract

In the present study, a new sorbent was fabricated from Palm kernel (PK) by dry thermochemical activation with NaOH and characterized by FTIR, X-ray diffraction, FE-SEM and BET, which was used for the Amoxicillin (AMX) sorption from aqueous solution. The influence of effective parameters such as pH, reaction time, adsorbent dosage, AMX concentration and ionic strength on the sorption efficacy of AMX removal were evaluated. The main functional groups on the surface of the magnetic activated carbon of Palm Kernel (MA-PK) were C-C, C-O, C[bond, double bond]O and hydroxyl groups. The specific surface of char, activated carbon Palm Kernel (AC-PK) and MA-PK were 4.3, 1648.8 and 1852.4 m²/g, respectively. The highest sorption of AMX (400 mg/L) was obtained by using 1 g/L of sorbent at solution pH of 5 after 60 min contact time, which corresponding to 98.77%. Non-linear and linear models of isotherms and kinetics models were studied. The data fitted well with Hill isotherm (R² = 0.987) and calculated maximum sorption capacity were 719.07 and 512.27 mg/g from Hill and Langmuir, respectively. A study of kinetics shows that the adsorption of AMX follows the Elovich model with R² = 0.9998. Based on the artificial neural network (ANN) modeling, the MA-PK dosage and contact time showed the most important parameters in the removal of AMX with relative importance of 36.5 and 25.7%, respectively. Lastly, the fabricated MA-PK was successfully used to remove the AMX from hospital wastewater.

Keywords: Adsorption; Agricultural waste; Artificial neural network; Isotherm and kinetic study; Wastewater treatment.