Spray Granulation Process Research Articles

Excipient quantitation and drug distribution during formulation optimization

An oral granules formulation experienced high drug content and increased variability when the process was scaled up from lab scale to clinical manufacturing scale. It was suspected that mannitol, due to its smaller particle size and lower density, was preferentially lost during the top spray granulation process, thereby causing active enrichment in the remaining granules. In order to troubleshoot the problem, rapidly evaluate solutions, and further optimize the formulation, a simple and rapid analytical technique was required. Since mannitol does not have a UV chromophore, conventional HPLC/UV analysis could not be used. Three alternative analytical techniques were evaluated in terms of ease of use, reproducibility, linear dynamic range and rapidity. The HPLC/RID (refractive index detector) and HPLC/ELSD (evaporative light scattering detector) provided rapid, reproducible alternate techniques to HPLC/UV, whereas LC/MS showed poor reproducibility. Analysis of the sieve samples of the granulations by HPLC/RID and HPLC/ELSD confirmed that poor active drug distribution was due to mannitol losses in the filter bag, as well as increased low size granules low in active drug content. The resultant formulation process was modified and a reduction in the initial air flow at start-up reduced losses of mannitol in the granulator filters.

Investigating the dynamic behaviour of fluidized bed spray granulation processes applying numerical simulation tools

Particulate processes involve different kinetic processes such as formation of nuclei, their subsequent growth and breakage. In addition, external product classification can play a significant role. A reliable prediction and thorough understanding of potential sources for instability is not only of scientific interest but also an important issue for better process design and process control. The application of these nonlinear dynamics is concerned with fluidized bed spray granulation. Self-sustained oscillations may rise in processes with external product classification. Therefore, focus of this work is on process stability which can be influenced by classifying, milling and recycling of particles and by the production of internal and external seeds. First a brief introduction to the model is given, that couples the particle population state with thermodynamic-, fluiddynamic- and granulation process for a unit with non-classifying product discharge and a screening and milling unit in the seed recycle and builds the core of the software package AVA ® FB Sim ® , used for the experiments.

A New Comprehensive Model and Simulation Package for Fluidized Bed Spray Granulation Processes

The model introduced in this paper makes it possible to calculate the expected product and bed material particle size distribution, as well as the dynamic behavior of process state variables. It was integrated in the developed software package FBSim and it could be shown that even a complex process like the fluidized bed spray granulation can be simulated in its dynamic behavior, if existing models are extended and coupled. The implemented model achieves a coupled solution of the population balance and heat and mass transfer equations.

Modeling the droplet deposition behavior on a single particle in fluidized bed spray granulation process

Fluidized bed spray granulation is used to produce coarse granular solid particles by spraying liquids in the form of fine droplets on a fluidized particle followed by drying in a stream of fluidizing air. The quality and thickness of granulation or coating on the particle should be uniform. The physical mechanism towards layering and the influence of process parameters towards granulation are not yet fully understood. In this work, the granulation process was carried out on a single spherical particle. Experiments were conducted to study the influence of process parameters, drying conditions, impact velocities and physical properties of sprayed solutions on the kinetics of granulation and on the morphology of the end product. The theoretical modeling of the process was attempted, relating collision and adhesion of sprayed droplets with particle growth for 20% NaCl solution using correlation modeling and artificial neural network techniques. The present paper reports and compares the results obtained by the two types of modeling. The results are useful in understanding the growth kinetics in spray-coating process.