A novel model for prediction solubility in binary solvent system

Abstract

A new seven parameter model for calculating and predicting solid solubility in binary solvent systems
has been presented as follows :
logxm=A0+A1/fA+A3(fA-FB)2+A4(FA+FB)3+A5(81-82/81)3+A5(V1-V2/V1)3
Where Xm is mole fraction solubility of solid, Ao-As denote the parameters or coefficients of the model, IA and fB are volume fractions of the solvents A and B in the absence of the solute in the mix'Iure,b1 and b2 represent the solubility parameters of the solvent mixture and solute, V1 and V2 are molar volumes of the solvent mixture and solute, respectively. I'A and I'B are given by I'A=0.96 fA+0.02 and I'B=0.96 Is+0.02 , V1 is calculated by means 01 V1=fA.VA+fBVS in which VA and Vs are molar volulmes of the neat solvents A and B. Also, ()1 is determined from : ()1 =fA,{)A+fB,{)B where OA and as represent the solubility parameters of the pure solvents. All the values of () and V can readily be calculated from the published group contribution tables available in literature. The coefficients of Ao-As can be easily estimated by means of a multiple regression analysis using an ordinary personal computer. The accuracy and prediction capability of the model were compared with those of six previous models i.e., log-linear, extended Hildebrand, polynomial statistical response mixture, nearlly ideal binary solvent (NIBS) , modified Wilson and double log-log, using data of 35 systems published in models in most cases. literature. The proposed model was superior to the mentioned