Solubility in binary solvent systems. IV. Prediction of naphthalene solubilities using the UNIFAC group contribution model

This work extends the UNIFAC group contribution method of solid-liquid equilibria to binary solvent mixtures, and compares its predictions to experimental solubilities for naphthalene in 16 different solvent mixtures. Deviations between experimental and calculated values are of the order of 10–20% for most solvent systems, and are comparable in magnitude to deviations noted in the pure solvents. The ability of the UNIFAC model to provide reasonable estimates of naphthalene solubilities based only on heat of fusion data and group contribution parameters suggests that the model may be useful in the area of drug design.     

Prediction of drug solubility in water-propylene glycol mixtures using Jouyban-Acree model

A trained version of the Jouyban-Acree model was presented to predict drug solubility in water-propylene glycol mixtures at various temperatures. The model is able to predict the solubility in various solubility units and requires the experimental solubility of a solute in mono-solvent systems. The mean percentage deviation (MPD) of predicted solubilities was computed to show the accuracy of the predicted data and 24% was found as the average MPD for 27 data sets studied. The proposed model enables the researchers to predict solubiliy in water-propylene glycol mixtures at various temperatures and reduces the number of required experimental data from five to two points.     

Proposition of group molar constants for sodium to calculate the partial solubility parameters of sodium salts using the van Krevelen group contribution method.

The aim of this study is to propose, for the first time, a set of group molar constants for sodium to calculate the partial solubility parameters of sodium salts. The values were estimated using the few experimental partial solubility parameters of acid/sodium salt series available either from the literature (benzoic acid/Na, ibuprofen acid/Na, diclofenac Na) or determined in this work (salicylic acid/Na, p-aminobenzoic acid/Na, diclofenac), the group contribution method of van Krevelen to calculate the partial parameters of the acids, and three reasonable hypothesis. The experimental method used is a modification of the extended Hansen approach based on a regression analysis of the solubility mole fraction of the drug lnX(2) against models including three- or four-partial solubility parameters of a series of pure solvents ranging from non-polar (heptane) to highly polar (water). The modified method combined with the four-parameter model provided the best results for both acids and sodium derivatives. The replacement of the acidic proton by sodium increased the dipolar and basic partial solubility parameters, whereas the dispersion parameter remained unaltered, thus increasing the overall total solubility parameter of the salt. The proposed group molar constants of sodium are consistent with the experimental results as sodium has a relatively low London dispersion molar constant (identical to that of -OH), a very high Keesom dipolar molar constant (identical to that of -NO(2), two times larger than that of -OH), and a very high hydrogen bonding molar constant (identical to that of -OH). The proposed values are: F((Na)d)=270 (J cm(3))(1/2) mol(-1); F((Na)p)=1030 (J cm(3))(1/2) mol(-1); U((Na)h)=17000 J mol(-1). Like the constants for the other groups, the group molar constants proposed for sodium are certainly not the exact values. However, they are believed to be a fair approximation of the impact of sodium on the partial solubility parameters and, therefore, can be used as such in the group contribution method of van Krevelen.     

Improved group contribution parameter set for the application of solubility parameters to melt extrusion

Hot-melt extrusion is gaining importance for the production of amorphous solid solutions; in parallel, predictive tools for estimating drug solubility in polymers are increasingly demanded. The Hansen solubility parameter (SP) approach is well acknowledged for its predictive power of the miscibility of liquids as well as the solubility of some amorphous solids in liquid solvents. By solely using the molecular structure, group contribution (GC) methods allow the calculation of Hansen SPs. The GC parameter sets available were derived from liquids and polymers which conflicts with the object of prediction, the solubility of solid drugs. The present study takes a step from the liquid based SPs toward their application to solid solutes. On the basis of published experimental Hansen SPs of solid drugs and excipients only, a new GC parameter set was developed. In comparison with established parameter sets by van Krevelen/Hoftyzer, Beerbower/Hansen, Breitkreutz and Stefanis/Panayiotou, the new GC parameter set provides the highest overall predictive power for solubility experiments (correlation coefficient r = −0.87 to −0.91) as well as for literature data on melt extrudates and casted films (r = −0.78 to −0.96).     

Prediction of the aqueous solubility: comparison of the general solubility equation and the method using an amended solvation energy relationship

An Amended Solvation Energy Relationship (ASER) was recently reported to successfully predict the aqueous solubilities of a set of 664 organic compounds. The average absolute error and root mean square error are 0.43 and 0.62 log units, respectively. When the General Solubility Equation (GSE) is applied to the same set of compounds, it gives an average absolute error of 0.45 log units and a root mean square error of 0.62 log units. These results are similar to those of the ASER method. The advantages and disadvantages of each method are discussed. It is shown that when the two methods agree with each other, they also agree with the experimentally determined values.     

In silico prediction of drug solubility in water-dioxane mixtures using the Jouyban-Acree model

A numerical method based on the Jouyban-Acree model was presented for prediction of drug solubility in water-dioxane mixtures at various temperatures. The method requires drug solubility in monosolvent systems, i.e. two data points for each temperature of interest. The mean percentage deviation (MPD) of predicted solubilities was calculated to show the accuracy of the predicted data and 27% was found as the average MPD for 36 data sets studied. The proposed numerical method reduced the number of required experimental data from five to two points and could also be extended to predict solubility at various temperatures.     

Prediction of drug solubility from structure

The aqueous solubility of a drug is an important factor affecting its bioavailability. Numerous computational methods have been developed for the prediction of aqueous solubility from a compound's structure. A review is provided of the methodology and quality of results for the most useful procedures including the model implemented in the QikProp program. Viable methods now exist for predictions with less than 1 log unit uncertainty, which is adequate for prescreening synthetic candidates or design of combinatorial libraries. Further progress with predictive methods would require an experimental database of highly accurate solubilities for a large, diverse collection of drug-like molecules.     

Prediction of aqueous solubility of a diverse set of compounds using quantitative structure-property relationships

"Fail early and fail fast" is the current paradigm that the pharmaceutical industry has adopted widely. Removing non-drug-like compounds from the drug discovery lifecycle in the early stages can lead to tremendous savings of resources. Thus, fast screening methods are needed to profile the large collection of synthesized and virtual libraries involved in the early stage. Solubility is one of the filters that are applied extensively to ensure that the compounds are reasonably soluble so that synthesis of the compounds and assay studies of pharmacokinetics and toxicity are feasible. To address this need, we have developed a fast quantitative structure-property relationship (QSPR) model for the prediction of aqueous solubility (at 298 K, unbuffered solution) from the molecular structures. Multiple linear regressions and genetic algorithms were used to develop the models. The model was based on a set of diverse compounds including small organic molecules and drug and drug-like species. The predicted solubility for the training and test sets agrees well with the experimental values. The coefficient of determination is R(2) = 0.84 for the training set of 775 compounds and the RMS error = 0.87. This model was validated on four sets of compounds. The RMS error for the 1665 compounds from the four validation data sets (including compounds from the Physician's Desk References and Comprehensive Medicinal Chemistry databases) is 1 log unit and the unsigned error is 0.77. This model does not require 3-D structure generation which is rather time-consuming. Using 2-D structure as input, this model is able to compute solubility for 90 000-700 000 compounds/h on a SGI Origin 2000 workstation. This kind of fast calculation allows the model to be used in data mining and screening of large synthesized or virtual libraries.     

Prediction of the acute toxicity (96-h LC50) of organic compounds to the fathead minnow (Pimephales promelas) using a group contribution method

A group contribution method has been developed to correlate the acute toxicity (96-h LC50) to the fathead minnow (Pimephales promelas) for 397 organic chemicals. Multilinear regression and computational neural networks (CNNs) were used for model building. The models were able to achieve a fairly good correlation of the data (r2 > 0.9). The linear model, which included four specific interaction terms, provided a rapid means of predicting the toxicity of a compound. The CNN model was able to yield virtually the same predictions with or without the four interaction terms that were included in the multilinear model.     

Molar solubility of felodipine in different aqueous systems

The molar solubility of felodipine in water was calculated and experimentally determined. A polyoxyethylene-polyoxypropylene block copolymer (Synperonic T304) and sodium lauryl sulfate as surfactants and ethanol as cosolvent were used as solubilizing agents. Significantly different molar solubilities were determined for felodipine according to the approach used to achieve solubilization. Such differences are the result of the different mechanisms of solubilization of the additives.     

Prediction of pharmaceutical solubility Via NRTL-SAC and COSMO-SAC

Solid phase solubility is a fundamental parameter in the design of crystallization processes. The development and optimization of crystallization processes requires screening of numerous solvent systems for which the solubility of the compound of interest has to be measured as a function of temperature and solvent composition. Tools that quickly estimate the solubility in different solvents can be very useful in the initial phases of the solvent system selection process. In this paper, we report our experience applying two thermodynamic models in the solubility estimation of pharmaceutical compounds: the NRTL-SAC method (Chen and Song, 2004, Ind Eng Chem Res 43: 8354) which provides a correlative and predictive model from limited solubility measurements, and the COSMO-SAC (Lin and Sandler, 2002, Ind Eng Chem Res 41: 899) method which predicts solubility from ab initio calculations. These theoretical methods, coupled with rapid experimental measurement for verification, provide a powerful solubility screening protocol for the development of crystallization processes.     

Novel liposome-based multicomponent systems for the protection of photolabile agents

A photosensitive drug (riboflavin) was entrapped as such or in the form of β- or γ-cyclodextrin complexes into the aqueous phase of multilamellar dehydration-rehydration vesicles (DRV liposomes) made of equimolar egg phosphatidylcholine or dipalmitoylphosphatidylcholine and cholesterol. Riboflavin-containing DRV were prepared in the absence or presence of one or more of the lipid-soluble UV absorbers oil red O, oxybenzone and dioxybenzone (entrapped into the lipid phase) and the water-soluble sulisobenzone (entrapped in the aqueous phase of liposomes together with riboflavin). In some experiments, lipid-soluble absorbers were supplemented with the antioxidant β-carotene. Entrapment values for free (41–47%) and complexed (19–23%) riboflavin were estimated fluorimetrically with additional data from NMR studies confirming that the complexes were entrapped as intact entities. Entrapment values for each of the UV light lipid-soluble absorbers (79–98%) and β-carotene (78 and 88%) were estimated by the use of the second-order derivative of their UV spectra to circumvent interference from overlapping absorption spectra of the other agents, when present. A number of conditions of entrapment were found to reduce values, for instance co-entrapment of sulisobenzone together with the vitamin in the case of riboflavin and, for all other materials, the absence (or reduced content) of cholesterol in DRV or certain variations in their manufacture. Exposure of a variety of riboflavin-containing DRV preparations to UV light revealed optimal protection with a formulation containing the γ-cyclodextrin complex of the vitamin, all three lipid-soluble light absorbers and β-carotene, increasing the half-life of riboflavin 266-fold. Results suggest that liposome-based multicomponent systems could be developed for the protection of photolabile agents in therapeutics and other uses.     

Models for calculating solubility in binary solvent systems

Two empirical models which express the relationship between the solute solubility and the concentration of one of the solvents in a binary solvent system are presented. The proposed models have been compared with previous similar models similar either in their original or modified forms from accuracy and predictability points of view using many experimental data taken from the literature. Both models were in some respects superior to the original and modified forms of the previous models. The modification of some of the previous models has improved the accuracy of the original models.