ABSTRACT
A better understanding of crystallization kinetics and the effect on drug product quality characteristics is needed to exploit the use of semi-crystalline polymers in pharmaceutical fused filament fabrication. Filaments were prepared from polycaprolactone or polyethylene oxide loaded with a crystallization inhibitor or inducer, which was either 10% (w/w) ibuprofen or theophylline. A design-of-experiments approach was conducted to investigate the effect of nozzle temperature, bed temperature and print speed on the printed tablets' microstructure and dissolution kinetics. Helium pycnometry derived porosity proved an ideal technique to capture significant distortions in the tablets' microstructure. On the other hand, terahertz time domain spectroscopy (THz-TDS) analysis proved valuable to investigate additional enclosed pores of the tablets' microstructure. The surface roughness was analyzed using optical coherence tomography, showing the importance of extensional viscosity for printed drug products. Drug release occurred via erosion for tablets consisting of polyethylene oxide, which partly reduced the effect of the inner microstructure on the drug release kinetics. An initial burst release effect was noted for polycaprolactone tablets, after which drug release continued via diffusion. Both the pore and crystalline microstructure were deemed essential to steer drug release. In conclusion, this research provided guidelines for material and process choice when a specific microstructure has to be constructed from semi-crystalline materials. In addition, non-destructive tests for the characterization of printed products were evaluated.
Subject(s)
Polyethylene Glycols , Polymers , Porosity , Drug Liberation , Tablets/chemistry , Polymers/chemistry , Technology, Pharmaceutical/methods , Printing, Three-Dimensional , SolubilityABSTRACT
An innovative continuous spin-freeze-drying technology for unit doses was recently developed. For this technology, a mechanistic primary drying model was developed allowing the calculation of the optimal dynamic drying trajectory for spin-frozen formulations. In this work, a model-based and experimentally verified comparison was made between conventional batch freeze-drying and spin-freeze-drying by analyzing the outputs (i.e., primary drying endpoint, optimal shelf temperature/power heater and product temperature profile) of both primary drying models. Input parameters such as dried product layer resistance (Rp) and heat input parameters (Kv,Ptot) were experimentally determined for both freeze-drying methods and compared. In addition, optimal dynamic process parameters were calculated for 3 model formulations by using both mechanistic models. Finally, model predictions were validated by measuring the product temperature and primary drying endpoint. It was observed that, when considering the same layer thickness, Rp was generally lower for continuous spin-frozen formulations compared to vials frozen in a conventional batch freeze-dryer. This observation contributes to the short primary drying times of spin-frozen formulations. In addition, as spin-freezing drastically increases the surface area of the product and lowers the dried layer thickness, drying times can be reduced even further while an excellent cake structure and appearance can still be obtained. The primary drying model for spin-frozen formulations proved to be equally accurate for the prediction of the primary drying endpoint and product temperature compared to the batch freeze-drying model.