Evaluation of thermodynamic contributions to extraction of medical devices by organic solvents as a sample preparation step in chemical characterization of medical devices

The thermodynamic contribution to extraction of medical devices by organic solvents as a first sample preparation step in chemical characterization studies is evaluated by Abraham's solvation parameter model using five representative materials (low density polyethylene or LDPE, silicone, polyurethane or PU, polyoxymethylene or POM, and polyacrylate or PA) and ten solvents (methanol, ethanol, isopropanol, acetonitrile, ethylene glycol, acetone, butanone, hexane, olive oil, and triolein). The Abraham's model is used to predict the material-solvent partition system coefficients by the corresponding partition system constants and representative extractables. The partition system constants are indirectly derived by a “thermodynamic circle conversion” method, based on material-water partition systems and solvent- water partition systems or material-air partition systems and solvent-air water partition systems. The material-solvent partition coefficient, PM/Solvent=CM/CSolvent, defined as the concentration in the material phase divided by the concentration in the solvent phase, is used as the solvent extraction strength. log(PM/Solvent) values are predicted for all material-solvent pairs using the representative extractables, mostly from Wayne state university experimental descriptor database (WSUEDD). The predictive log(PM/Solvent) values of material-solvent pairs are considered as the upper bound, indicating the significance of partitioning effect in solvent extraction. The calculation results using water-based or air-based partition systems are also compared. The predictive results are discussed in relation to the solvent-material interaction or swelling as well. Several conclusions can be drawn from this study. First, the predictive consistency of two conversion systems (water-based or air-based) is established, indicating the accuracy and robustness of Abraham's model. Second, the predicted partition coefficients are confirmed by available experimental values (LDPE and silicone), and the predicted solvent extraction strengths are supported by available experimental extraction data. Third, the kinetic effect, rather than the thermodynamic effect, is the dominant extractables release process in sample preparation step of chemical characterization studies. The solvent selection in these studies should be optimized based mainly on the diffusional kinetics and solvent-material interactions (swelling effect). Fourth, acetone and butanone can be the general-purpose solvent for the extraction of all materials, thereby eliminating the need for three solvents in chemical characterization studies. Finally, Abraham's solvation parameter model is demonstrated as an invaluable tool in understanding and differentiating the solvent extraction processes.