Linear solvation energy relationship for the adsorption of synthetic organic compounds on single-walled carbon nanotubes in water

Han Ding,Chao Chen,Xiaojian Zhang

DOI: https://doi.org/10.1080/1062936X.2015.1132764

IF: 3.681

2016-01-01

SAR and QSAR in Environmental Research

Abstract:The linear solvation energy relationship (LSER) was applied to predict the adsorption coefficient (K) of synthetic organic compounds (SOCs) on single-walled carbon nanotubes (SWCNTs). A total of 40 log K values were used to develop and validate the LSER model. The adsorption data for 34 SOCs were collected from 13 published articles and the other six were obtained in our experiment. The optimal model composed of four descriptors was developed by a stepwise multiple linear regression (MLR) method. The adjusted r(2) (r(adj)(2)) and root mean square error (RMSE) were 0.84 and 0.49, respectively, indicating good fitness. The leave-one-out cross-validation Q(2) ( [GRAPHICS] ) was 0.79, suggesting the robustness of the model was satisfactory. The external Q(2) ( [GRAPHICS] ) and RMSE (RMSEext) were 0.72 and 0.50, respectively, showing the model's strong predictive ability. Hydrogen bond donating interaction (bB) and cavity formation and dispersion interactions (vV) stood out as the two most influential factors controlling the adsorption of SOCs onto SWCNTs. The equilibrium concentration would affect the fitness and predictive ability of the model, while the coefficients varied slightly.