Prediction of solvent removal profile and effect on properties for peptide-loaded PLGA microspheres prepared by solvent extraction/evaporation method

Using a predictive mathematical model, several important extrinsic process variables were varied to simulate the process dynamics of microsphere formation. These included the composition profile in the dispersed phase, the solvent concentration profile in the continuous phase and the solvent removal profile in the dispersed phase. By superimposing the composition profile in the dispersed phase with the phase transition boundary, the progression of phase transition in microsphere formation can be evaluated. Low dispersed phase/continuous phase ratio, high continuous phase-addition rate, high temperature, high heating rate and high initial polymer concentration in the dispersed phase contributed to enhanced solvent removal. The higher solvent removal led to a heterogeneous composition distribution in the dispersed phase and the early cross-over of the gelation point (viscous boundary) of the periphery region which initiates the onset of solidification in this region. These phenomena resulted in an increasing pore size, lower surface area, denser periphery, higher residual solvent and slower drug release. In addition, the progress toward the glassy boundary may also play a major role in the ultimate solvent residual. Slow solvent removal gave rise to a homogenous distribution of the components in the dispersed phase due to the delay of hardening. The extrinsic manageable parameters could be varied during microsphere formation to obtain the desired rate of solvent removal as well as the desired microsphere properties. The mathematical model was used to simulate such conditions to facilitate the experimental design for the desired microsphere properties.