QSPR for predicting the hydrophile-lipophile balance (HLB) of non-ionic surfactants

The hydrophile-lipophile balance (HLB) value provides an important reference for evaluating the performance behaviors of surfactants. In this work, a quantitative structure-property relationship (QSPR) model was established to predict HLB values of non-ionic surfactants based on the norm descriptor concept. The results showed that the calculated HLB values of 237 non-ionic surfactants agreed well with the experimental values with the squared correlation coefficient (R2) for the whole dataset of 0.9901 and the average absolute relative deviation (AARD) of 2.93 %. The R2 and AARD of the training set and the testing set are 0.9901 (Rtraining2) and 0.9900 (Rtesting2), 2.92 % and 3.47 %, respectively. The cross-validation results, Y-randomized test, mean absolute error (MAE) test and application domain (AD) analysis suggested that this QSPR model performs well in accuracy, robustness and reliability. These results demonstrated that this model is accurate and stable, and further validated that the norm descriptor concept is suitable for describing the HLB values of non-ionic surfactants.