Solubility In Binary Solvent Mixtures Research Articles

Towards the prediction of drug solubility in binary solvent mixtures at various temperatures using machine learning

Drug solubility is an important parameter in the drug development process, yet it is often tedious and challenging to measure, especially for expensive drugs or those available in small quantities. To alleviate these challenges, machine learning (ML) has been applied to predict drug solubility as an alternative approach. However, the majority of existing ML research has focused on the predictions of aqueous solubility and/or solubility at specific temperatures, which restricts the model applicability in pharmaceutical development. To bridge this gap, we compiled a dataset of 27,000 solubility datapoints, including solubility of small molecules measured in a range of binary solvent mixtures under various temperatures. Next, a panel of ML models were trained on this dataset with their hyperparameters tuned using Bayesian optimization. The resulting top-performing models, both gradient boosted decision trees (light gradient boosting machine and extreme gradient boosting), achieved mean absolute errors (MAE) of 0.33 for LogS (S in g/100 g) on the holdout set. These models were further validated through a prospective study, wherein the solubility of four drug molecules were predicted by the models and then validated with in-house solubility experiments. This prospective study demonstrated that the models accurately predicted the solubility of solutes in specific binary solvent mixtures under different temperatures, especially for drugs whose features closely align within the solutes in the dataset (MAE < 0.5 for LogS). To support future research and facilitate advancements in the field, we have made the dataset and code openly available.Scientific contributionOur research advances the state-of-the-art in predicting solubility for small molecules by leveraging ML and a uniquely comprehensive dataset. Unlike existing ML studies that predominantly focus on solubility in aqueous solvents at fixed temperatures, our work enables prediction of drug solubility in a variety of binary solvent mixtures over a broad temperature range, providing practical insights on the modeling of solubility for realistic pharmaceutical applications. These advancements along with the open access dataset and code support significant steps in the drug development process including new molecule discovery, drug analysis and formulation.Graphical

Solid-liquid equilibrium solubility prediction of sulfanilamide in four binary solvent mixtures by using pure solvents solubility data from 278.15 to 318.15 K with the Abraham solvation parameter model, Yalkowsky Log-Linear and extended log-linear solubility thermodynamic models

The industrial application of sulfanilamide solubility in binary solvent mixtures is a critical factor in drug development, assessment, as well as future improvement. The current work eliminates the substantial experimental work, as well as the ability to apply improved and upgraded computed models that include solute physicochemical properties and activity coefficients. The predicted solid–liquid equilibrium data of sulfanilamide in acetone containing binary solvent mixtures at temperatures ranging from 278.15 to 318.15 K were calculated by applying the combined version of Jouyban-Acree model with Abraham solute parameter model, combined version of Jouyban-Acree model with Abraham solvation parameter model, combined version of Jouyban-Acree model with Hansen solubility parameter model, the Yalkowsky Log-Linear model, Extended log-linear model. In comparison to previously published binary solvent solubility data, this paper provides predicted binary solubility data using only mono-solvent solubility data as well as from the computed models. From the solubility trained model the %ARD can be ranked as the combined version of combined version of the Jouyban-Acree model with Abraham Solute parameter model (19.47 %) < the Jouyban-Acree model with Hansen solubility parameter model (23.76 %) < the Yalkowsky model (50.54 %) < the combined version of Jouyban-Acree model with Abraham solvation parameter model (68.83 %) < the Extended log-linear model (615.78 %). The combined version of Jouyban-Acree model with Abraham Solute parameter model giving most reliable predicted solubility results compared to other computed models. The most important outcome of this research will be immensely beneficial in calculating binary solubility data utilizing only mono-solvent solubility data as well as from computed models. The binary solubility predictions were calculated from already reported mono-solvent solubility data from experimentally determined, Hildebrand-Scatchard equation-1 and Hildebrand-Scatchard equation-2.

Non-electrostatic energies as a metric for prediction of deferasirox solubility in binary solvent mixtures: Polarized Continuum Model tactic

The aim of this work is to enhance deferasirox (DFX) solubility by employing the co-solvency approach. Drug solubility was studied in six different binary solvent mixtures including acetone (1) + water, acetone (1) + methanol, acetone (1) + acetonitrile, glycerol (1) + methanol, glycerol (1) + ethanol, glycerol (1) + water at 298.15 ± 0.05 K. Molecular modeling is employed to study the solvent-solvent interactions involved in the co-solvency data using the density functional theory and polarizable Continuum Model (DFT/PCM). The accuracy of the DFT/PCM theory is evaluated by comparison of its prediction values with the experimental solubility data. These results lead to the conclusion that, for acetone + water, acetone + methanol, acetonitrile + acetone and water + glycerol systems non-electrostatic effects are more important in solvent-solvent interactions. While for other solvent systems electrostatic energies are dominant. This work constitutes a major advance in the application of non-electrostatic energiesin DFT/PCM calculation and prediction of drug solubility pattern in some binary solvent mixtures. Semi-empirical models are used to determine the predictability power of the experimental co-solvency data. The combined nearly ideal binary solvent/Redlich-Kister (CNIBS/R-K) and the modified Wilson models were able to predict the experimental solubility data with MRD value of less than 35%.

Measurement and correlation of solubility and thermodynamic properties of dicumyl peroxide (DCP) in pure and binary solvents

The solid-liquid equilibrium data of dicumyl peroxide (DCP) in seven pure solvents and ethanol-water binary system and DMF-water binary system were measured from 274.15 K to 302.15 K by a gravimetric method under atmospheric pressure. The results show that the solid-liquid equilibrium solubility of DCP rapid increased with augmented temperature in different types of pure solvent while the solubility in binary solvent mixtures increases with the increasing of the fraction of ethanol and DMF. The measured data in pure solvents were correlated with several thermodynamic models including the modified Apelblat model, λh model and NRTL model. All the experimental results were agreement with the fitted values. Moreover, the NRTL model was adopted to calculate the enthalpy, entropy and Gibbs energy change during the dissolution process. And for the binary solvent system, the CNIBS/R-K model was used to correlate the solubility data. It was found that the correlated results by the NRTL model agreed best with the experimental data. Besides the results indicate that the mixing process of dicumyl peroxide in the experimental solvents was spontaneous and entropy-driven.

Solubility and thermodynamic parameters of 5-Fluoro-2-oxindole in nine pure solvents and binary solvent mixtures at T = (278.15–323.15) K

In this research, we employed a static equilibrium method to measure the solubility of 5-Fluoro-2-oxindole in nine pure solvents (n-hexane, n-propanol, tetrahydrofuran, isopropanol, methanol, ethyl acetate, N,N-dimethylformamide, acetonitrile, acetone) and three binary solvents (n-hexane+ tetrahydrofuran, ethyl acetate + tetrahydrofuran, acetonitrile + tetrahydrofuran) from 278.15 K to 323.15 K. The results demonstrate that as the temperature increases, the solubility of 5-Fluoro-2-oxindole increases in nine solvents selected. The modified Apelblat model and the λh model were adopted to correlate the solubility in pure solvents; meanwhile, we employed the CNIBS/R–K model and the Jouyban–Acree model to correlate the solubility in binary solvent mixtures. The maximum RAD values of modified Apelblat model and λh model are 0.64% and 0.71%; respectively; the maximum RMSD values are 0.076% and 0.058%. The corresponding maximum RAD values of the CNIBS/R–K model and the Jouyban–Acree model are 1.27% and 2.22%. In general, these equations emerge an excellent fitting effect.

Experimental Determination and Thermodynamic Models for Solid–Liquid Equilibrium of 4-(4-Aminophenoxy)-N-methylpyridine-2-carboxamide in Pure and Binary Solvent Mixtures for T = (278.15–328.15) K

In this study, the solubility of 4-(4-aminophenoxy)-N-methylpyridine-2-carboxamide (APMC) in pure tetrahydrofuran, ethyl acetate, acetonitrile, n-butanol, n-propanol, ethanol, methanol, and three binary solvent mixtures (ethyl acetate + methanol, ethyl acetate + ethanol, and ethyl acetate + n-butanol) was first experimentally determined within the temperature range from (278.15 to 328.15) K under atmosphere pressure by the gravimetric method. The results reveal that the solubility of APMC increases with an increase in temperature in all of the solvents selected. The modified Apelblat equation and the Buchowski–Ksiazaczak λh equation were used to correlate the experimental solubility in pure solvents while the modified Apelblat equation, the CNIBS/R–K model, and the Jouyban–Acree model were applied to correlate the solubility in binary solvent mixtures. All of the models could be well applied, and the modified Apelblat equation stood out to be more suitable with a higher accuracy compared to that of the ot...

Solubility of Benzoin in Six Monosolvents and in Some Binary Solvent Mixtures at Various Temperatures

The solubility of benzoin in monosolvents (acetone, ethyl acetate, methanol, ethanol, 1-propanol, and 1-butanol) and binary solvent mixtures (ethyl acetate + methanol, ethyl acetate + ethanol) was measured using UV–vis spectroscopy at temperatures ranging from 283.15 K to 323.15 K. It can be seen from the data that the solubility of benzoin increases expectedly as temperature increases in a given solvent or solvent mixture, the solubility in acetone is maximum among six monosolvents which could be well explained by the existence of strong H-bonds, rather than the “like dissolves like” rule. In binary solvent mixtures, the solubility reaches maximum when the mole fraction of methanol is 0.1 in ethyl acetate + methanol mixed solvents, while the maximum exhibits at 0.2 of mole fraction of ethanol in ethyl acetate + ethanol. The solubility parameter was interpreted as the cosolvency of benzoin solubility in binary solvent mixtures. The solubility data were correlated by modified Apelbalt equation, CNIBS/R-K e...

Solubility and Thermodynamic Behavior of Veramoss in Different Pure Solvents and {Ethanol + Water} Mixtures from T = 283.15 to 333.15 K

The solubility of veramoss in n-butyl alcohol, ethanol, isopropanol, ethyl acetate, acetone, tetrahydrofuran, isoamyl alcohol, n-amyl alcohol, n-propyl alcohol, and DMF pure solvents and various {ethanol + water} mixtures at T = 283.15–333.15 K and atmospheric pressure was measured by using the dynamic method. The experiment solubility data showed that the solubility of veramoss increased with increasing temperature in all solvents selected and decreased with the rise of mole fraction of water in the binary solvent mixtures of {ethanol + water}. The modified van’t Hoff equation and the Apelblat model were successfully used to correlate the experimental solubility in pure solvents while the modified Apelblat model, the van’t Hoff equation, and the CNIBS/R model were applied to correlate the solubility in binary solvent mixtures, and all of the models provided good fitting to the data. Furthermore, the thermodynamic parameters (dissolving enthalpy, dissolving entropy, and Gibbs free energy) of veramoss in d...

Determination and thermodynamic modelling for 2-methyl-6-nitroaniline solubility in binary solvent mixtures of ethyl acetate + (methanol, ethanol, n-propanol and isopropanol)

The solubility of 2-methyl-6-nitroaniline in binary mixed solvents of (ethyl acetate+methanol), (ethyl acetate+ethanol), (ethyl acetate+n-propanol) and (ethyl acetate+isopropanol) was determined experimentally by using the isothermal dissolution method within the temperature range from (278.15–313.15)K under atmosphere pressure. The solubility of 2-methyl-6-nitroaniline increased with increasing temperature and mass fraction of ethyl acetate in each binary system. At the same temperature and mass fraction of ethyl acetate, the mole fraction solubility of 2-methyl-6-nitroaniline was greater in (ethyl acetate+n-propanol) than in the other three mixed solvents. The achieved solubility values were correlated by employing Jouyban-Acree model, van’t Hoff-Jouyban-Acree model, Apelblat-Jouyban-Acree model, Ma model, Sun model and CNIBS/R-K model. The values of relative average deviations (RAD) and root-mean-square deviations (RMSD) were no greater than 1.79% and 20.56×10−4 for the six models. In general, the Jouyban-Acree model proved to provide better representation of the experimental solubility. Furthermore, the dissolution enthalpies of the dissolution process were calculated. Positive values of dissolution enthalpy showed that the dissolution process of 2-methyl-6-nitroaniline in these mixed solvents was endothermic. The experimental solubility results and models presented in the present work are essential in the practical process for production and purification of 2-methyl-6-nitroaniline.

Thermodynamic models for determination of 3-chloro-N-phenylphthalimide solubility in binary solvent mixtures of (acetone, ethyl acetate or 1,4-dioxane + methanol)

The solubility of 3-chloro-N-phenylphthalimide in binary mixed solvents of (acetone+methanol, ethyl acetate+methanol and 1,4-dioxane+methanol) were determined experimentally by using the isothermal dissolution equilibrium method within the temperature range from (288.15 to 323.15)K under atmosphere pressure. For the binary systems of (acetone+methanol) and (1,4-dioxane+methanol), the solubility of 3-chloro-N-phenylphthalimide increased with increasing temperature and mass fraction of acetone or 1,4-dioxane; and for the (ethyl acetate+methanol) system, at a given composition of ethyl acetate, the solubility of 3-chloro-N-phenylphthalimide increased with an increase in temperature; nevertheless at the same temperature, they increased at first and then decreased with increasing mass fraction of 1,4-dioxane. At the same temperature and mass fraction of acetone, ethyl acetate or 1,4-dioxane, the solubility of 3-chloro-N-phenylphthalimide was greater in (1,4-dioxane+methanol) than in the other two mixed solvents. The solubility values were correlated by employing the Jouyban–Acree model, van’t Hoff–Jouyban–Acree model, Apelblat–Jouyban–Acree model, Ma model, and Sun model. On the whole, the Ma model and Sun model were proven to provide good representation of the experimental solubility results. Furthermore, the dissolution enthalpies of the dissolution process were calculated. The dissolution process of 3-chloro-N-phenylphthalimide in these mixed solvents is endothermic. The experimental solubility and the models in this study could be helpful in purifying 3-chloro-N-phenylphthalimide.

Measurement and correlation of solubility of cefmenoxime hydrochloride in pure solvents and binary solvent mixtures

The solubility of cefmenoxime hydrochloride in pure solvents and binary solvent mixtures was measured at temperatures from (283.15 to 313.15)K by using the UV spectroscopic method. The results reveal that the solubility of cefmenoxime hydrochloride increases with increasing temperature in all solvent selected. The solubility of cefmenoxime hydrochloride reaches its maximum value when the mole fraction of isopropanol is 0.2 in the binary solvent mixtures of (isopropanol+water). The modified Apelblat equation and the NRTL model were successfully used to correlate the experimental solubility in pure solvents while the modified Apelblat equation, the CNIBS/R–K model and the Jouyban–Acree model were applied to correlate the solubility in binary solvent mixtures. In addition, the mixing thermodynamic properties of cefmenoxime hydrochloride in different solvents were also calculated based on the NRTL model and experimental solubility data.

Determination and correlation of solubility of spironolactone form II in pure solvents and binary solvent mixtures

The solubility data of spironolactone form II in six pure solvents and binary solvent mixtures of ethyl acetate and methanol were measured over the temperature range from (278.85 to 317.75)K by using a dynamic method under atmospheric pressure. The results show that the solubility of spironolactone form II in pure solvents increases with increasing temperature while the solubility in binary solvent mixtures increases with the increasing of the fraction of ethyl acetate. This phenomenon is well explained by using relative dielectric constants of the solvents. The solubility data of spironolactone form II in pure solvents were well correlated by the modified Apelblat equation, the Wilson model and the NRTL model while the solubility of spironolactone form II in binary solvent mixtures were correlated by the modified Apelblat equation and the CNIBS/R-K model. Furthermore, the thermodynamic properties of the dissolution process of spironolactone form II were also determined by using the van’t Hoff equation.

Solubility in binary solvent mixtures: Pyrene dissolved in alcohol + toluene mixtures

The solubilities of pyrene in mixed solvents (methanol+toluene, ethanol+toluene and 2-propanol+toluene) were determined at temperatures between 283K and 333K by a gravimetric method. The experimental data were correlated with the Apelblat model. The calculated values of Apelblat model were found to show a fine representation of the experimental data.

Naphthalene Solubility in Binary Solvent Mixtures of 2,2,4-Trimethylpentane + Alcohols at 298.15 K

The solubility of naphthalene in binary mixtures of 2,2,4-trimethylpentane + C1−C4 alcohols was determined at 298.15 K. The solubility increased with the addition of the alcohols to 2,2,4-trimethylpentane, reached the maximum values, and then decreased with further increase of alcohol concentrations. Results of these measurements were used to evaluate the prediction capability of a previously developed QSPR models employing the solubility data in monosolvents, and the overall mean deviation (OMD) of the models was varied between (6.0 and 19.1) %. Using ab initio prediction methods, the OMDs varied between (27.6 and 30.6) %.

Mathematical Representation of Solubility of Electrolytes in Binary Solvent Mixtures Using Jouyban-Acree Model

Applicability of a solution model for calculating salt's solubility in binary solvent mixtures was shown. The accuracy of the proposed model was evaluated by computing mean percentage deviation (MPD) employing available solubility data of electrolytes in binary solvents at various temperatures from the literature. The overall MPD (+/-S.D.) for correlation of solubility data was 4.7+/-5.1% and for prediction using model trained by a minimum number of experimental data points was 10.5+/-12.5%.

Solubility in Binary Solvent Mixtures: Anthracene Dissolved in Alcohol + Carbon Tetrachloride Mixtures at 298.2 K

Article discussing solubility in binary solvent mixtures and anthracene dissolved in alcohol + carbon tetrachloride mixtures at 298.2 K.

Solubility in Binary Solvent Mixtures: Anthracene Dissolved in Alcohol + 2-Methyl-1-butanol Mixtures at 298.2 K

Experimental solubilities are reported for anthracene dissolved in seven binary alcohol + 2-methyl-1-butanol solvent mixtures at 25 °C. The alcohol cosolvents studied were 1-propanol, 2-propanol, 1-butanol, 2-butanol, 2-methyl-1-propanol, 3-methyl-1-butanol, and 2-pentanol. Results of these measurements are used to test a mathematical representation based upon the combined nearly ideal binary solvent (NIBS)/Redlich−Kister equation. For the seven systems studied, the combined NIBS/Redlich−Kister equation was found to mathematically describe the experimental data to within an overall average absolute deviation of approximately ±0.4%.

Solubility in Binary Solvent Mixtures: Pyrene Dissolved in Alcohol + Acetonitrile Mixtures at 299.2 K

Experimental solubilities are reported for pyrene dissolved in seven binary alcohol + acetonitrile solvent mixtures at 26 °C. The alcohol cosolvents studied were 1-propanol, 2-propanol, 1-butanol, 2-butanol, 2-methyl-1-propanol, 3-methyl-1-butanol, and 2-pentanol. Results of these measurements are used to test a mathematical representation based upon the Combined Nearly Ideal Binary Solvent (NIBS)/Redlich−Kister equation. For the seven systems studied, the Combined NIBS/Redlich−Kister equation was found to mathematically describe the experimental data to within an overall average absolute deviation of approximately ±1.0%.

Solubility in Binary Solvent Mixtures: Anthracene Dissolved in Alcohol + Acetonitrile Mixtures at 298.2 K

Experimental solubilities are reported for anthracene dissolved in six binary alcohol + acetonitrile solvent mixtures at 25 °C. The alcohol cosolvents studied were methanol, ethanol, 1-pentanol, 2-pentanol, 2-methyl-1-pentanol, and 4-methyl-2-pentanol. Results of these measurements are used to test a mathematical representation based upon the combined nearly ideal binary solvent (NIBS)/Redlich−Kister equation. For the six systems studied, the combined NIBS/Redlich−Kister equation was found to mathematically describe the experimental data to within an overall average absolute deviation of approximately ±0.7%.

Comparison of various cosolvency models for calculating solute solubility in water–cosolvent mixtures

Previously published cosolvency models are critically evaluated in terms of their ability to mathematically correlate solute solubility in binary solvent mixtures as a function of solvent composition. Computational results show that the accuracy of the models is improved by increasing the number of curve-fit parameters. However, the curve-fit parameters of several models are limited. The combined nearly ideal binary solvent/Redlich–Kister, CNIBS/R–K, was found to be the best solution model in terms of its ability to describe the experimental solubility in mixed solvents. Also presented is an extension of the mixture response surface model. The extension was found to improve the correlational ability of the original model.