Upscaling and in-line process monitoring via spectroscopic techniques of ethylene vinyl acetate hot-melt extruded formulations

Abstract

The aim of the present work was to evaluate drug release and quality of EVA/drug matrices at different PEO 7M concentrations (5 and 15%), manufactured using two different hot-melt extruders: a lab-scale mini extruder and a pilot-scale extruder. The process parameters used on both extruders (temperature and screw speed) and drug release from the matrices were compared. On the lab-scale extruder all formulations were extruded at 90 °C, whereas on the pilot-scale extruder the temperature of the die was adjusted to 100 °C in order to achieve a constant pressure at the extrusion die, hence constant material flow through the die to yield smooth extrudates. Screw speed was also adjusted from 60 rpm (lab-scale extruder) to 90 rpm (pilot-scale extruder) in order to obtain a balance between feeding rate and screw speed. Drug release from the obtained matrices on both extruders was also assessed. Despite the differences in diameter (diameter of 2 and 3mm for the lab-scale extruder and pilot-scale extruder, respectively), temperature and screw speed, drug release per surface area was similar. DSC analysis of a formulation [EVA40/MPT (50/50, w/w) with 5% PEO] indicated small changes in its solid state after extrusion on both extruders: drug crystallinity was reduced by max. 20%, PEO recrystallized after cooling and EVA remained semi-crystalline. Extrusion experiments on the pilot-scale extruder of EVA/MPT, 50/50 (w/w) formulations were also monitored in-line using Raman and NIR spectroscopy in order to evaluate the material behavior at a molecular level in the extrusion barrel as function of the process settings (extrusion temperature: 90, 110 and 140 °C; screw speed: 90 and 110 rpm). At 90 and 110 °C the crystallinity of the drug was reduced, but the majority of MPT remained in its crystalline state as specific peaks in the Raman spectra of the drug became broader. These differences were accentuated when extrusion was performed at 140 °C as the drug completely melted. Peak shifts to lower frequencies [(CO) groups of the drug and (CH(3)COO) groups of EVA] were registered at all extrusion temperatures, with maximum effect at 140 °C indicating molecular interactions. Increasing the screw speed did not result in peak shifts of Raman spectra. NIR confirmed these observations and showed an additional peak in the spectra characteristic of (OH) bounds.