Premium
Mercury removal from water using deep eutectic solvents‐functionalized multi walled carbon nanotubes: Nonlinear autoregressive network with an exogenous input neural network approach
Author(s) -
Fiyadh Seef S.,
AlSaadi Mohammed A.,
Binti Jaafar Wan Z.,
AlOmar Mohamed K.,
Fayaed Sabah S.,
Hama Ako R.,
Hin Lai S.,
ElShafie Ahmed
Publication year - 2019
Publication title -
environmental progress and sustainable energy
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.495
H-Index - 66
eISSN - 1944-7450
pISSN - 1944-7442
DOI - 10.1002/ep.13261
Subject(s) - nonlinear autoregressive exogenous model , mean squared error , autoregressive model , adsorption , approximation error , biological system , nonlinear system , carbon nanotube , artificial neural network , materials science , root mean square , mathematics , chemistry , statistics , computer science , nanotechnology , physics , machine learning , quantum mechanics , biology
This work presents the experimental and modeling process for mercury ions removal from water using functionalized multi‐walled carbon nanotube as adsorbent. The modeling procedure has been carried out using nonlinear autoregressive network with an exogenous input (NARX) neural network modeling technique is used for modeling the adsorbent's adsorption capacity using different parameters based on experimental data. The effect of different parameters including mercury ions concentration, pH, amount of adsorbent dosage, and contact time is studied. Three kinetics models such as intraparticle diffusion, pseudo first‐order, and pseudo second order are applied using the experimental and predicted outputs, the pseudo second order was the best to describe. A sensitivity study is conducted using different parameters. Various indicators are applied to examine the accuracy and efficiency of the NARX model such are mean square error (MSE), root mean square error, relative root mean square error, mean absolute percentage error, relative error (RE), and coefficient of determination ( R 2 ). The value of the maximum RE was 3.49%, the R 2 was 0.9998, and the MSE was 4.28 × 10 −6 . Based on the used indicators, the NARX model was capable to predict the adsorbent's adsorption capacity by comparing the NARX model outputs to the experimental results.