Prediction of some important physical properties of sulfur compounds using quantitative structure–properties relationships

Prediction of critical properties of sulfur-containing compounds: New QSPR models

In this study, new models have been proposed for the prediction of different critical properties (critical temperature (T_C), critical pressure (P_C), critical volume (V_C), and acentric factor (ω)) of the sulfur-containing compounds based on quantitative structure-property relationship (QSPR). An extensive data set containing experimental data of over 130 different sulfur-containing compounds was employed. Enhanced Replacement Method (ERM) was applied for subset variable selection. Based on ERM selected descriptors, two different models, including linear model and genetic programming (GP) based non-linear model have been proposed for each critical property. The predicted values of each target were in good agreement with the experimental data. For GP-based models, the values of the coefficient of determination (R²) were 0.936, 0.976, 0.990, and 0.917 for T_C, P_C, V_C, and ω, respectively. After revisiting the available QSPR models, it was found that the domain of applicability of new models has been expanded.

Solvation Entropy Made Simple

The entropies of molecules in solution are often calculated using gas phase formulas. It is assumed that, because implicit solvation models are fitted to reproduce free energies, this is sufficient for modeling reactions in solution. However, this procedure exaggerates entropic effects in processes that change molecularity. Here, computationally efficient (i.e., having similar cost as gas phase entropy calculations) approximations for determining solvation entropy are proposed to address this issue. The S_ω, S_ϵ, and S _ϵα models are nonempirical and rely only on physical arguments and elementary properties of the medium (e.g., density and relative permittivity). For all three methods, average errors as compared to experiment are within chemical accuracy for 110 solvation entropies, 11 activation entropies in solution, and 32 vaporization enthalpies. The models also make predictions regarding microscopic and bulk properties of liquids which prove to be accurate. These results imply that Δ H_sol and Δ S_sol can be described separately and with less reliance on parametrization by a combination of the methods presented here with existing, reparametrized, implicit solvation models.

Comparison between Multi-Linear- and Radial-Basis-Function-Neural-Network-Based QSPR Models for The Prediction of The Critical Temperature, Critical Pressure and Acentric Factor of Organic Compounds

Critical properties and acentric factor are widely used in phase equilibrium calculations but are difficult to evaluate with high accuracy for many organic compounds. Quantitative Structure-Property Relationship (QSPR) models are a powerful tool to establish accurate correlation between molecular properties and chemical structure. QSPR multi-linear (MLR) and radial basis-function-neural-network (RBFNN) models have been developed to predict the critical temperature, critical pressure and acentric factor of a database of 306 organic compounds. RBFNN models provided better data correlation and higher predictive capability (an AAD% of 0.92–2.0% for training and 1.7–4.8% for validation sets) than MLR models (an AAD% of 3.2–8.7% for training and 6.2–12.2% for validation sets). The RMSE of the RBFNN models was 20–30% of the MLR ones. The correlation and predictive performances of the models for critical temperature were higher than those for critical pressure and acentric factor, which was the most difficult property to predict. However, the RBFNN model for the acentric factor resulted in the lowest RMSE with respect to previous literature. The close relationship between the three properties resulted from the selected molecular descriptors, which are mostly related to molecular electronic charge distribution or polar interactions between molecules. QSPR correlations were compared with the most frequently used group-contribution methods over the same database of compounds: although the MLR models provided comparable results, the RBFNN ones resulted in significantly higher performance.

QSPR models for predicting generator-column-derived octanol/water and octanol/air partition coefficients of polychlorinated biphenyls

Octanol/water (K(OW)) and octanol/air (K(OA)) partition coefficients are two important physicochemical properties of organic substances. In current practice, K(OW) and K(OA) values of some polychlorinated biphenyls (PCBs) are measured using generator column method. Quantitative structure-property relationship (QSPR) models can serve as a valuable alternative method of replacing or reducing experimental steps in the determination of K(OW) and K(OA). In this paper, two different methods, i.e., multiple linear regression based on dragon descriptors and hologram quantitative structure-activity relationship, were used to predict generator-column-derived log K(OW) and log K(OA) values of PCBs. The predictive ability of the developed models was validated using a test set, and the performances of all generated models were compared with those of three previously reported models. All results indicated that the proposed models were robust and satisfactory and can thus be used as alternative models for the rapid assessment of the K(OW) and K(OA) of PCBs.

A QSPR model for prediction of diffusion coefficient of non-electrolyte organic compounds in air at ambient condition

Evaluation of diffusion coefficients of pure compounds in air is of great interest for many diverse industrial and air quality control applications. In this communication, a QSPR method is applied to predict the molecular diffusivity of chemical compounds in air at 298.15K and atmospheric pressure. Four thousand five hundred and seventy nine organic compounds from broad spectrum of chemical families have been investigated to propose a comprehensive and predictive model. The final model is derived by Genetic Function Approximation (GFA) and contains five descriptors. Using this dedicated model, we obtain satisfactory results quantified by the following statistical results: Squared Correlation Coefficient=0.9723, Standard Deviation Error=0.003 and Average Absolute Relative Deviation=0.3% for the predicted properties from existing experimental values.

An accurate model for prediction of autoignition temperature of pure compounds

Accurate prediction of pure compounds autoignition temperature (AIT) is of great importance. In this study, the Artificial Neural Network-Group Contribution (ANN-GC) method is applied to evaluate the AIT of pure compounds. 1025 pure compounds from various chemical families are investigated to propose a comprehensive and predictive model. The obtained results show the squared correlation coefficient of 0.984, root mean square error of 15.44K, and average percent error of 1.6% for the experimental values.

A QSPR model for estimation of lower flammability limit temperature of pure compounds based on molecular structure

In this study, a quantitative structure-property relationship was presented to estimate lower flammability limit temperature (LFLT) of pure compounds. This relationship is a multi-linear equation and has six parameters. These chemical structure-based parameters were selected from 1664 molecular-based parameters by genetic algorithm multivariate linear regression (GA-MLR). Since 1171 compounds were used to develop this equation, the model can be used to estimate the LFLT of a wide range of pure compounds.

Current mathematical methods used in QSAR/QSPR studies

This paper gives an overview of the mathematical methods currently used in quantitative structure-activity/property relationship (QASR/QSPR) studies. Recently, the mathematical methods applied to the regression of QASR/QSPR models are developing very fast, and new methods, such as Gene Expression Programming (GEP), Project Pursuit Regression (PPR) and Local Lazy Regression (LLR) have appeared on the QASR/QSPR stage. At the same time, the earlier methods, including Multiple Linear Regression (MLR), Partial Least Squares (PLS), Neural Networks (NN), Support Vector Machine (SVM) and so on, are being upgraded to improve their performance in QASR/QSPR studies. These new and upgraded methods and algorithms are described in detail, and their advantages and disadvantages are evaluated and discussed, to show their application potential in QASR/QSPR studies in the future.

Prediction of the Standard Enthalpy of Formation of Pure Compounds Using Molecular Structure

A predictive approach has been presented to calculate the standard enthalpy of formation of pure compounds based on a quantitative structure–property relationship technique. A large number (1692) of pure compounds were used in this study. A genetic algorithm based on multivariate linear regression was used to subset variable selection. Using the selected molecular descriptors an optimized feed forward neural network was presented to predict the ΔHfo of pure compounds.

Estimation of molecular diffusivity of pure chemicals in water: a quantitative structure-property relationship study

A quantitative structure-property relationship (QSPR) study was performed to predict the molecular diffusivity of pure chemicals in water. A genetic-algorithm-based multivariate linear regression (GA-MLR) was applied to select the most statistically effective molecular descriptors for modelling the molecular diffusivity of pure chemicals in water. Based on the selected molecular descriptors, a three-layer feed forward neural network (FFNN) was constructed to predict the property. The obtained results showed that the FFNN was able to predict the molecular diffusivity of pure chemicals in water.