Effect of Plasticizer On Glass Transition Temperature of Amorphous Pharmaceuticals - A Computational Study

 

PURPOSE

The objectives were: (i) to develop a molecular dynamics (MD) based technique for predicting glass transition temperature (T_g) and (ii) to computationally evaluate the effect of plasticizer (water) on T_g of a model amorphous compound (sucrose, SUC). 

METHODS

The periodic in-silico models of amorphous SUC containing 0%, 3% and 5%w/w water were constructed at 298K.  COMPASS forcefield was utilized for the MD simulations.  The cells were energy minimized (steepest descent, conjugate gradient), and then equilibrated (2ns) using canonical (NVT) and isothermal-isobaric (NPT) ensembles at 440K.  Each system was quench cooled from 440K to 265K at intervals of 5K.  The specific volumes (V) were computed using the cell densities and plotted as a function of the corresponding temperature (T).  Three independent simulations for each system were performed to improve sampling statistics.

RESULTS

The computed specific density of the amorphous SUC cells at 298K was in agreement with experimentally determined value.  A gradual increase in density of the amorphous cells, along with corresponding decrease in specific volume, was noted as each system was quenched.  As the system approached glassy state, a discontinuity in V-T curve was observed. This characteristic ‘kink' represents the transition from less stable ‘rubbery state' to a kinetically stable ‘glassy state'.  Statistical regression analysis of the data corresponding to the glassy and the rubbery region identified the T_g to be 367K.  The RSD for data points corresponding to the rubbery phase was greater than that for the glassy state, attributable to greater molecular mobility in the rubbery state.  Increasing the water content progressively decreased the system T_g.  The MD computed T_g values for amorphous SUC containing 3% and 5% w/w water, were 352K and 343K respectively.

CONCLUSIONS

The T_g of amorphous SUC was successfully computed by means of full atomistic based MD simulations. Enhanced mobility in the rubbery state was also manifested as increased scatter of data points in the high temperature regimes.  The simulated plasticization effect corroborated well with theoretical model predictions and reported experimental data.  The technique can thus provide a molecular level insight in to the physical stability of amorphous formulations during product development.