Purpose The purpose of this investigation is to develop modeling and simulation tools to conduct in-silico design of delivery systems for pharmaceuticals, in particular tablets. A basic component is the systematic correlation of end product quantities (e.g. dissolution profile) with the physical characteristics of the tablet (e.g. spatial distribution of particles). In this paper we describe a methodology to determine optimal active distribution for a prescribed dissolution profile. This methodology integrates physically-based solvers that track the effects of fluid penetration with nonlinear optimization techniques. This methodology opens the possibility of aiding in the design process of tablets with desired or required performance.
Methods The computational tool initially simulates tablets with a predefined distribution of active nano-particles on a three dimensional fixed grid. A steady-state concentration is used to develop a localized flux for each of the nano-particles, and determine the amount of time it would take for each of the nano-particles to dissolve completely. All of the nano-particle volumes are reduced accordingly in steps and particles below a critical volume are removed from the simulation, until all of the nano-particles have dissolved completely. The outputs of these simulations are then compared to the desired output and the differences in composition to determine a rate of change. This rate of change is then used to develop a new particle distribution whose output is closer to the desired output. This process is repeated until the desired output is reached to within an acceptable tolerance.
Results The simulation produces profiles for active release, water concentration and surface fluxes at every time step. The program is easily adaptable to multiple grid sizes, active nano-particle distribution concentrations, and diffusion properties through several global constants. The backsolving algorithm is capable of dealing with particle distribution profiles of varying complexity.
Conclusions This program provides a good basis for future extensions and demonstrates feasibility of the method. The process can be adapted into larger simulations incorporating a moving fluid boundary layer, variable tablet geometry, and composite particle systems. The backsolving algorithm is removable from the dissolution program and is easily adapted to future dissolution simulations.