Multi-parameter experiments and modeling for nitrate sorption to quaternary ammonium-functionalized poly(amidoamine) dendrimers in aqueous solutions

초록

The aim of this study was to investigate nitrate sorption onto quaternary ammonium-functionalized poly(amidoamine) dendrimer generation 2.0 (q-PAMAM-G2). The physicochemical characteristics of the q-PAMAM-G2 were examined using field emission scanning electron microscopy, thermogravimetric analysis, electrophoretic light scattering spectrophotometry, elemental analysis, etc. Single-parameter sorption tests were performed under batch conditions to examine the nitrate sorption characteristics of the q-PAMAM-G2. According to a pH experiment, the q-PAMAM-G2 maintained a relatively stable nitrate sorption capacity at pH 2 to 10. X-ray photoelectron spectrometry analysis indicated a new peak for the NH4NO3 bond after nitrate sorption, confirming nitrate sorption to quaternary ammonium groups on the surface of the q-PAMAM-G2 through anion exchange. The kinetic data were best fit with the pseudo-second-order model, whereas the equilibrium data fitted well with the Langmuir isotherm, suggesting that nitrate sorption onto the q-PAMAM-G2 occurred through chemisorption. Multi-parameter sorption experiments (n = 32) were conducted with three input variables (adsorbent dose, initial nitrate concentration, and solution pH) and one output variable (nitrate removal rate). In response surface methodology (RSM) modeling, a quartic regression model (fourth-order polynomial equation) was developed to predict nitrate sorption onto the q-PAMAM-G2. In the artificial neural network (ANN) modeling, a model with a structure of 3:10:1 was derived to describe the nitrate sorption data. Additional multi-parameter experiments for nitrate sorption onto the q-PAMAM-G2 (n = 8) were conducted to further evaluate the developed models. The developed ANN model (R-2 = 0.872) predicted better than did the RSM model (R-2 = 0.790) for the additional multi-parameter experimental data.

키워드