Scale-up of an oil/water cream containing 40% diethylene glycol monoethyl ether.

The purpose of this study was to scale up an oil/water (o/w) cream formulation containing 40% diethylene glycol monoethyl ether (DGME), developed via 300-g laboratory batches in a 2(5-2) fractional factorial design, to 7-kg batch sizes in a Brogli-10 homogenizer. The o/w cream was manufactured via a standard phase-inversion process in the Brogli-10 homogenizer. Partitioning studies of DGME were conducted in test tubes at ambient temperature and after 24 hr at 70 degrees C in a convection oven. Phase height was measured by vernier calipers. Microscopy studies of excipients with and without treatment with water or a DGME/water mixture were conducted with a Nikon microscope after equilibration at 35 degrees C for 24 hr. During creation of the 7-kg pilot-scale batches, congealed material was observed between the sweep agitation blade and the discharge port, where the Brogli-10 homogenizer is not temperature jacketed. Factors that increased the amount of congealed material were higher temperatures during primary emulsification and longer cooling times. Partitioning studies revealed that DGME resides in the aqueous external phase of this formulation. Microscopy studies revealed that DGME in the external phase of this cream has a profound impact on the solubility of certain solid, waxy excipients (e.g., cetyl alcohol and polyoxyethylene-2-stearyl ether) at 35 degrees C. From this study, it appears that DGME resides in the external phase of the o/w cream. During manufacturing, it is hypothesized that the presence of DGME in the external phase alters the solubility of certain solid, waxy excipients in the formula such that they no longer primarily reside in the internal oil phase. On cooling, these materials precipitate or congeal in the external phase. The fractional factorial experimental design at the 300-g laboratory scale did not predict the issues encountered during scale-up. Differences between laboratory scale and pilot plant scale that explain why this phenomenon was not seen during laboratory scale are differences in cooling times, nonjacketed or "cold spots" in the Brogli-10 homogenizer, and a low proportion of congealed material in relation to the total batch size (< 1.5%).