Graphene oxide-cyanuric acid nanocomposite as a novel adsorbent for highly efficient solid phase extraction of Pb2+ followed by electrothermal atomic absorption spectrometry; statistical, soft computing and mechanistic efforts

In this research, graphene oxide-cyanuric acid (GO-CA) nanocomposite as an efficient and novel adsorbent was used for solid phase extraction of Pb2+ followed by electrothermal atomic absorption spectrometry (ETAAS). The synthesised adsorbent was characterised by the Fourier transform-infrared spectrophotometry (FT-IR), field emission scanning electron microscopy (FE-SEM) and Energy dispersive X-Ray spectroscopy (EDX). In order to optimise the parameters including pH of sample solution, amounts of adsorbent, extraction and desorption times; the response surface methodology based on central composite design (RSM-CCD) was used. Under the optimum conditions, the calibration curve was linear in the range of 0.15-3.0 mu g L-1 Pb2+. Also, the limit of detection (LOD) and the relative standard deviation were 0.021 mu g L-1 (n=7) and 3.1 % (seven replicate analysis of 1 mu g L-1 Pb2+), respectively. To evaluate the adsorption mechanism of Pb2+ onto the GO-CA nanocomposite, two parameter (Langmuir and Freundlich) and three parameter isotherms were evaluated and based on the obtained results, the adsorption of Pb2+ onto the GO-CA nanocomposite governed by the Freundlich isotherm with a the maximum adsorption capacities of 333.0 mg g(-1). According to the kinetic models including the Pseudo First Order (PFO), Pseudo Second Order (PSO), Intra-particle diffusion, Elovich and Boyd models; adsorption of Pb2+ followed by the second-order kinetic model and film diffusion is the rate controlling step. Moreover, based on the geometric computations, the adsorption and desorption processes don't have any interfering together during the contact time. In the following, Artificial Neural Network (ANN) and Random Forest algorithm (RFA) are employed for prediction of adsorption performance based on effective parameters such as pH, amounts of adsorbent, extraction and desorption times. The outcomes of soft computing (ANN and RFA) illustrated acceptable accuracy (R-2>0.92) for estimation and prediction of extraction recovery of Pb2+