Investigation on Key Parameters of Common Pharmaceutical Blending Unit Operations using Discrete Element Method
Yijie Gao, Sree Chalasani, Bhavishya Mittal
Takeda Pharmaceuticals International Co.
Keywords: high shear blending, Discrete Element Method, binding index, material properties, process parameters, full-factorial DoE, surface area ratio
Blending is a commonly used unit operation in many solid oral dosage drug products. In the case of high-shear wet granulation (HSWG), blending is considered as an essential pre-step in order to break drug substance agglomerates and is the key to reduce segregation tendency especially when cohesive drug substance is involved [1-6]. The purpose of this study is to simulate the high-shear blending processes using Discrete Element Modeling (DEM) in order to understand the influence of several material properties and process parameters on the breakage of drug substance agglomerates and their subsequent mixing with chemically inert excipients.
High-shear blending was numerically investigated with drug substance and excipients simulated by DEM particles using the commercially available software Star CCM+ v10.2 (CD-Adapco). The computations were done using dual CPUs, Intel® Xeon® E5-2640 @ 2.50 GHz (24 processors). Simulation platforms were designed in Star CCM+ to mimic the particle properties with specific particle size and flowability. The geometry of the Diosna High Shear Granulator 1-liter bowl was applied and imported into the software. Excipient particles were modeled using spheres with 5 mm (D), and drug substance particles were modeled using cylindrical rods of a smaller size with 1 mm (D) x 4 mm (L). During the simulations, a new index, called binding index, was proposed to characterize the tendency of how drug substance agglomerates break and how drug substance and excipient particles bind together. The binding index is defined as the total number of drug substance and excipient particle interactions divided by the total number of drug substance particles. The influence of 7 factors was studied: excipient particle size (4 – 6 mm); particle true density (0.55 – 2.20 g/mL); surface friction coefficient (0.1 – 0.6); work of cohesion (0 – 1.0 J/m2); drug load percentage (0.55% – 3.23%); impeller speed (100 – 300 rpm); and chopper speed (0 – 500 rpm). The surface friction and work of cohesion value ranges were set based on real material flowability measurements using a Jenike & Johanson ring shear tester. A total of 24 high-shear blending simulations were performed, and the binding index growth curves were monitored. The binding index growth curve was then fitted using an exponential model. The steady state binding index and the characteristic binding time were obtained from the fitting as the high-shear blending indices.
The high-shear blending binding indices were plotted to show the influence of the 7 factors. Based on the results, work of cohesion, particle size, and the drug load percentage are the most influential factors on the indices during high-shear blending in the current scope of the investigation. In order to understand how to select an excipient for the high-shear blending process, a further full-factorial DoE with 3 factors was performed: particle size, drug load percentage, and surface friction. The results confirmed that drug load percentage and excipient particle diameter affected the steady state binding index (p-value = 0.0049 and p-value = 0.0106, respectively), and that drug load percentage affected the characteristic binding time (p-value = 0.0065).
From this research, the surface area ratio between drug substance and excipient particles is the most significant factor of the high-shear blending unit operation. This research demonstrates the advantage of the DEM technology on providing practical guidance on the selection of excipient properties and process parameters in the formulation design and process development of solid dose drug products.
1. Nguyen, D., A. Rasmuson, I. Niklasson Björn, and K. Thalberg, CFD simulation of transient particle mixing in a high shear mixer. Powder Technology, 2014. 258(0): p. 324-330.
2. Nakamura, H., H. Fujii, and S. Watano, Scale-up of high shear mixer-granulator based on discrete element analysis. Powder Technology, 2013. 236(0): p. 149-156.
3. Levin, M., Wet Granulation: End-Point Determination and Scale-Up, in Encyclopedia of Pharmaceutical Technology, Third Edition. 2013, Taylor & Francis. p. 4078-4098.
4. Bridson, R.H., P.T. Robbins, Y. Chen, D. Westerman, C.R. Gillham, T.C. Roche, and J.P.K. Seville, The effects of high shear blending on α-lactose monohydrate. International Journal of Pharmaceutics, 2007. 339(1–2): p. 84-90.
5. Badawy, S.F. and M. Hussain, Effect of starting material particle size on its agglomeration behavior in high shear wet granulation. AAPS PharmSciTech, 2004. 5(3): p. 16-22.
6. Litster, J.D., K.P. Hapgood, J.N. Michaels, A. Sims, M. Roberts, and S.K. Kameneni, Scale-up of mixer granulators for effective liquid distribution. Powder Technology, 2002. 124(3): p. 272-280.