Prediction of liquid viscosity for organic compounds by a quantitative structure-property relationship

Abstract

The liquid viscosity of 361 organic compounds containing C, H, N, O, S and/or halogens was investigated using a quantitative structure-property relationship (QSPR) approach. A five-descriptor equation was obtained with a squared correlation coefficient (R 2 ) of 0.854 and a standard error (S) of 0.22 log units. The data set was divided into three subsets and each of these was cross-validated by the model obtained from the complementary two subsets; the result indicates a high statistical robustness of the overall model. The average absolute error S 2 of 0.0494 log units is small compared with the overall range of eta of 1490 mPas.

Application of a General Computer Algorithm Based on the Group-Additivity Method for the Calculation of Two Molecular Descriptors at Both Ends of Dilution: Liquid Viscosity and Activity Coefficient in Water at Infinite Dilution

“…In the present study the viscosity coefficients have been transformed into their decimal logarithm and entered into the group-additivity calculation as log(η). The main sources of experimental viscosity data have been the collective papers of Suzuki et al [5,7] and Katritzky et al [6], supplemented by more recently published experimental results for alkanes [11,12,13,14], haloalkanes [15,16], alkanols [17,18,19,20], alkylamines [21,22,23,24], aminoalcohols [25,26,27], ethers [28,29], aminoethers [30], acetals [31], ketones [32], esters [33,34,35,36,37,38,39,40,41,42,43], hydroxyesters [44,45], carbonate esters [46], and amides [47,48,49,50,51,52]. Beyond these, experimental data have been added for compounds with atom groups that have not yet been represented in the parameters table: phosphoric acid esters [53,54,55], phosphoric acid amides [56], siloxanes [57] and in particular ionic liquids [32,58,59,60,61,62,…”

“…Earlier attempts to predict the liquid viscosity coefficient of organic compounds have been developed on a statistical mechanics model based on the square well intermolecular potential [3], or have been carried out applying multiple linear regression and artificial neural network modelling methods using a limited number of descriptors as input [4,5], or are based on a quantitative structure-property relationship (QSPR) approach using a five-descriptor equation [6], or use a combination of partial least-square and QSPR technique starting with 18 mostly experimental parameters, finally ending with a model with nine descriptors [7]. …”

A General Treatment of Solubility. 3. Principal Component Analysis (PCA) of the Solubilities of Diverse Solutes in Diverse Solvents

“…the BMLR algorithm [25][26][27] for the development of the QSPR models have been added and the pool of calculated descriptors has been increased with up to 40 hydrogen-bonding descriptors.…”

“…The second descriptor is the hydrogenbonding donor charged surface area, HDCA(1), defined as a Where N is the number of data points, n is the number of parameters in the model, R 2 and R 2 cv are the square of the correlation coefficient, and cross-validation correlation coefficient, respectively, s 2 represent the standard deviation, and F is the Fisher's criterion. 27 The combination of the two descriptors (D17 and D7) evidently represents adequately the intermolecular forces that influence the solubility process. The gravitation index (D17) is related to the dispersion and cavity-formation effects in liquids.…”

“…29 Two-parameter QSPR models for liquid viscosity (log η) also included the same two descriptors (G I and HDCA-2) giving correlation coefficient R 2 ) 0.79 using 337 30 and R 2 ) 0.81 using 361 diverse organic molecules, respectively. 27,31 For the transposed small matrix, SM2 (where the solvents are observations and the solutes are variables), the first four factors together cover only 65.95% of the information. The measure of the quality of prediction, Q 2 , decreases after the fourth PC that indicates that introducing a higher number of PC into the PCA model is not appropriate (see Table 3).…”