Prediction of aqueous solubility of alcohols by molecular approach
عنوان دوره: 1
گروه مهندسی شیمی، دانشگاه صنعتی بیرجند، بیرجند، ایران
In this work, two new models by group contribution method (GC) and quantitative structure-property relationship (QSPR) were developed to prediction the aqueous solubility of alcohols based on theoretically derived molecular parameters. To develop the GC model a total of 53 substructures or functional groups was used. An efficient approach combining modified particle swarm optimization (MPSO) and multiple linear regression (MLR) was used for selecting a subset of relevant descriptors and building the optimal QSPR model. The proposed models were assessed by external validation method. The squared correlation coefficient values of 0.900 for the GC model and 0.897 for the QSPR model show good predictive ability of both models. The selected descriptors by MPSO-MLR suggest that size and hydrophobicity of alcohols are the most important factors affecting their aqueous solubility.
Aqueous solubility؛ Alcohol؛ Group contribution method (GC)؛ Quantitative structure property relationship (QSPR)؛ Modified particle swarm optimization (MPSO)؛ Multiple linear regression (MLR)