(91b) DEM Simulation of the Spheronization Process with Pharmaceutical Pellets Using a Contact Model Developed Based on Experimental Results

Introduction
Oral drug delivery is the most popular way of drug delivery because of the ease of ingestion, the avoidance of pain and also the lower production costs [1]. In the oral administration there are basically two different types, single unit dosage forms and multiple unit dosage forms. Conventional tablets belong to the single unit dosage forms. Smaller multi-particulate systems of pellets filled into capsules or compressed into tablets which disintegrate into the original pellets when taken, can be assigned to multiple dose unit forms [2]. Numerous advantages of multiple unit dosage forms compared to single unit dosage forms can be found in [3, 4]. The pellets used in multiple unit dosage forms are characterized by a spherical shape as well as a narrow size distribution. Beside some other pelletization techniques, a combined extrusion and spheronization process is widely used to produce such pellets [5, 6]. A spheronizer consists of a stationary cylindrical wall and a rotating disk with a structured surface. The rounding of the cylindrical extrudates of the extrusion step is influenced by various overlapping mechanisms. Those mechanisms are highly dependent on the particle dynamics in the spheronizer. For this reason, simulations of the particle dynamics with the Discrete Element Method (DEM) are a useful tool to get detailed information of what happens during the pellet rounding and to describe the spheronisation process. In the DEM, each particle is described separately via the equations of motion for translation and rotation, taking into account the interactions with other particles and apparatus walls. For those interactions, a model is needed which describes the deformation behavior of the pellets. In this study a contact model was derived from different experiments with wet spheronized pellets. Moreover, the results of the simulations using the developed contact model were compared with a simpler linear model.

Materials and Experiments
For the experiments pellets were produced with the combined extrusion and spheronization process. A mixture consisting of microcrystalline cellulose (Vivapur 102, JRS Pharma, Germany) with a mass fraction of wi = 0.2 g/g and α-lactose monohydrate (Granulac 200, Meggle, Germany) with wi = 0.8 g/g was extruded (ZSE 27 MAXX, Leistritz, Germany) with water at 200 rpm. Afterwards, 300 g of the extrudates were spheronized in a lab scale spheronizer (R250, Gabler, Germany) with a rotational speed of the friction plate of 750 rpm for 5 minutes. The resulting spheronized pellets have a desired moisture content of about 40%.
To obtain the force-displacement behavior of the pellets, uniaxial compression tests with the Texture Analyser (TA.XTplus, Stable Micro Systems, UK) were performed. Since the moisture content of the pellets has a significant impact on the mechanical properties, drying of the pellets during the experiments must be avoided. Therefore a climatic chamber was used to control the temperature and the moisture during the compression tests. The cyclic deformation of pellets was studied at different force levels and stress rates. In our previous work [7], it was shown that the energy absorption during compression and impact tests may differ due to the different stressing rates. Therefore, impact tests at different impact velocities were performed in addition to the compression tests to determine the coefficient of restitution as shown in [7].

Contact Model and Simulation
In our previous works [8, 9], the force-displacement behavior was approximated with linear relations according to the model of Walton&Braun [10, 11]. In [12], different approaches to model plastic material behavior based on the work of Tomas and Luding can be found. In this contribution, a new contact model is presented. The force-displacement relationship during loading and reloading is still modelled using a linear function, however the unloading stage is approximated with a power law to further increase the accordance of model and experiment. Moreover, the change in stiffness for reloading, because of plastic deformation and consolidation in the contact area, is considered, using the flattening of the particle. The energy dissipation in the model is due the different stiffnesses for loading and unloading. Thus the unloading dependents on the coefficient of restitution, which is a function of the impact velocity and the loading cycle.
The DEM simulations of the particle dynamics in a spheronization process, using the enhanced contact model, were compared against simulations based on the linear contact model regarding computing time and deviation of the results. The collision rates and forces were in particular focus. Distributions of the collision rates and the time averaged collision forces in the spheronizer were obtained.

References
[1] A. Fasano, Journal of Pharmaceutical Sciences 1998, 87 (11), 1351 – 1356. DOI: 10.1021/js980076h.
[2] J. M. Newton, International Journal of Pharmaceutics 2010, 395 (1-2), 2 – 8. DOI: 10.1016/j.ijpharm.2010.04.047.
[3] V. R. Sirisha K., K. Vijaya Sri, K. Suresh, G. Kamalakar Reddy, N. Devanna, IJPSR 2013, 4 (6). DOI: 10.13040/IJPSR.0975-8232.4(6).2145-58.
[4] H. Bechgaard, G. H. Nielsen, Drug Development and Industrial Pharmacy 2008, 4 (1), 53 – 67. DOI: 10.3109/03639047809055639.
[5] S. Muley, T. Nandgude, S. Poddar, Asian Journal of Pharmaceutical Sciences 2016, 11 (6), 684 – 699. DOI: 10.1016/j.ajps.2016.08.001.
[6] N. Shinde, N. Aloorkar, A. Kulkarni, B. Bangar, S. Sulake, P. Kumbhar, Asian Journal of Pharmaceutical Sciences 2014 (4), 38 – 47.
[7] S. Antonyuk, S. Heinrich, J. Tomas, N. G. Deen, M. S. van Buijtenen, J. A. M. Kuipers, Granular Matter 2010, 12 (1), 15 – 47. DOI: 10.1007/s10035-009-0161-3.
[8] D. Weis, M. Niesing, M. Thommes, S. Antonyuk, Chemie Ingenieur Technik 2017, 89 (8), 1083 – 1091. DOI: 10.1002/cite.201600154.
[9] D. Weis, M. Niesing, M. Thommes, S. Antonyuk, F. Radjai, S. Nezamabadi, S. Luding, J. Y. Delenne, EPJ Web Conf. 2017, 140, 15005. DOI: 10.1051/epjconf/201714015005.
[10] O. R. Walton, R. L. Braun, Acta Mechanica 1986, 63 (1-4), 73 – 86. DOI: 10.1007/BF01182541.
[11] O. R. Walton, R. L. Braun, Journal of Rheology 1986, 30 (5), 949 – 980. DOI: 10.1122/1.549893.
[12] R. Tykhoniuk, J. Tomas, S. Luding, M. Kappl, L. Heim, H.-J. Butt, Chemical Engineering Science 2007, 62 (11), 2843 – 2864. DOI: 10.1016/j.ces.2007.02.027.