(W1030-05-32) Use of DEM to Troubleshoot Aesthetic Defects in Pharmaceutical Tablets

Purpose: Pharmaceutically elegant tablets are an expectation from health care providers and consumers for the solid oral dosage forms. Periodically, non-aesthetically pleasing defects can be present in the finished product caused by non-robust tablet cores or the design and operation of the coating process. If the defect level doesn’t meet current standards the batch can be considered unfit for use and then must be sorted or possibly discarded. The purpose of this study is to simulate and analyze the design of the tablet core and coating process to gain a better understanding of the generation of tablet defect in an effort to help eliminate them from the finished product.

Methods: As physical trial-and-error method is most often impractical due to material constraints and other resources, use of virtual methods become necessary in such cases. This study employs Discrete Element Method (DEM) using the software product Altair® EDEM™ to understand the defects produced in such a coating process on a production scale. Traditionally, DEM has been conducted with multispherical particles, but since the tablet shape plays a vital role in the location of the defect, multi-faceted polyhedral shaped particles are considered to represent the physical shape of the tablet. An industrial scale coater operation is simulated using a typical production batch size of tablet cores, Figure 1. A Lagrangian based approach is used to record all the contacts of each of the ~200,000 tablets in the system during thirty seconds of operation. A total of 98 million contacts were analyzed in this time period and the forces acting on each individual vertex of the tablet during the process, due to collision with another tablet or with the coating pan itself, were calculated. These force values on the vertices are then compared to determine the region of the tablet with maximum propensity for breakage or defect formation.

Results: The value of the forces acting on each of the particles and the location of the impact for each of the contact was recorded at every 0.1 seconds. Due to this frequency of data recording, it is possible that the peak forces may not be recorded if the peak occurs between data save periods. However, since the process was run for twenty-five seconds over the steady state, the study assumes that the highest recorded forces are indicative of impact events that lead to tablet defects. The contacts with the highest ten force values were analyzed, and it revealed that most of the peak impact forces were created by lifting of the tablet by the coater baffle to the top of the coater and being dropped down by gravity to the surface of the coater. One of the other top ten force values were also due to tablet impact on another tablet during mixing of the tablets, but those forces were comparatively less significant than the tablets impacting the coating pan even though the majority of the tablet-to-tablet contacts would be observed during the mixing of the tablets in the avalanche region of the coating pan. A heat map, Figure 3, of the impact forces on the tablets were plotted to show the distribution of these forces on the tablets and the region/type of defects on the tablets. Most of the high impact forces on the tablets were observed on their edges, Figure 3, edge defects would seem to be more common compared to the face defects based on the force analysis. The tablet regions were divided into two groups – edge and face, and the forces on each vertex were summed up across the time intervals for both groups. The simulation found a roughly 70/30 split between edge defects and face defects. To validate the defects caused by the high force impacts, a statistical inspection of actual coated tablets was conducted for production batches. The results of these inspections showed that the amount of edge defects, i.e., edge chipping, was much greater than any deformity of the tablet in the face region of the tablet. It was noted in the inspection that, while limited to no physical damage of the tablet face was observed due to the high forces, if the coating conditions were on the wetter side picking defects were observed, Figure 3. This would not be unexpected due to the number of tablet-to-tablet interactions that would be occurring during the coating process while the tablets are avalanching in the spray zone of the coating pan.

Conclusion: The physical validation done for the simulation gives confidence in the simulation and motivates to pursue multi-faceted DEM technology in identifying tablet defects in the manufacturing process before the physical production of the tablets. These simulations can be used to aid tablet design, and to modify the equipment or manufacturing process to minimize the defective tablets created during processing resulting in a pharmaceutically elegant solid oral dosage form.

Tablet and coating pan designs used in DEM

Tablet movement for ten highest forces observed during the coating operation

Defects comparison with simulation (left) and observed on tablet edges (chipping, middle) and tablet faces (picking, right)