(84d) Role of the Ejection Process in the Capping of Pharmaceutical Biconvex Tablet after Compaction: A FEM Study

Tablet is the most popular pharmaceutical form. Its manufacturing using die compaction has been used for more than a century. Nevertheless, classical problems encountered during compaction are still to be solved. Among them, capping corresponds to the separation of the top layer of the tablet and can be observed just at the ejection from the die or some days after. A large number of articles have been written on the subject to establish correlations between powder/process parameters and capping tendency of the tablet. Nevertheless the problem is still not fully understood.

One of the known parameter influencing capping is the shape of the tablet. It is well known that biconvex tablets are more prone to capping than flat-faced ones. The reason is the mechanism that makes the tablet break. Various publications have described this mechanism. Forty years ago, Hiestand et al. explained that during relaxation, the material in the cup could expand radially whereas the main body could not explain radially. This promotes the development of large shear stresses that can lead to capping. More recently, FEM modelling was used to illustrate this mechanism. It confirms that the elastic recovery of the tablet during unloading promotes the development of a high shear stress at the limit between the land and the cup of the tablet. This shear stress is likely to be responsible of the capping of biconvex tablets which can thus be considered as a failure in shear.

Considering this mechanism, if the compaction is applied symmetrically, the shear stress considered is the same on both side of the tablet. Capping should thus appear on both side of the tablet. Nevertheless it is well known that, sometimes, capping can be asymmetric, i.e. failure occurs only on one side of the tablet. When this phenomenon happens, capping always occurs on the side that is ejected first (i.e. the “upper side” on most of the machines). This dissymmetry is not explained by the mechanism proposed above.

In this presentation we will study how the ejection part of the compaction cycle could be responsible of this asymmetrical failure. In fact the ejection process is clearly asymmetric. It is thus interesting to study how the stresses inside the tablet vary during the ejection process, especially the shear stress at the limit between the land and the cup of the tablet.

Experimental characterization of the stresses inside the tablet during the process is difficult. But, as already demonstrated in a lot of studies, numerical modelling using finite element method (FEM) makes it possible to access the stresses inside the tablet during the whole compaction cycle. In this work, FEM modelling, based on the Drucker Prager Cap model for the powder behavior, was used to study the influence of the ejection process on the stresses inside the tablet and to explain the role of the ejection process in the capping of biconvex tablet. Afterwards, influence of some process/product parameters on the capping tendency of biconvex tablet were discussed, based on the phenomena found in the simulations.

In a previous publication, we focused on the elastic recovery of a biconvex tablet during decompression using FEM. One of the findings was the fact that the contact between the punch (upper and lower punches) and the tablet was not lost at the same time on all the points of the surface. The contact was lost gradually from the land to the center of the cup, due to the elastic recovery of the cup of the tablet. A logical consequence of this typical elastic recovery is that the radius of curvature of the tablet at the end of the decompression is lower than the radius of curvature of the punch.

This elastic recovery has also consequences on the first part of the ejection process. During this part the lower punch moves up in the die. As the curvature of the tablet is higher than the curvature of the punch (i.e. the radius of curvature is smaller), the contact between the punch and the tablet will begin at the center of the cup. Because of the frictions between the tablet and the die and because of the residual die wall pressure (RDP), a certain force (ejection force) is needed to move the tablet upward. Until this force is reached, the band of the tablet will not move even if the punch is moving up. When the lower punch moves up, the contact area between the punch and the tablet increases without any movement of the tablet’s band. The consequence is a deformation of the tablet during this part.

Thanks to FEM modelling, it is possible to study the influence of this deformation on the stress distribution inside the tablet. Attention was focused on the shear stress at the limit between the land and the cup on the upper and lower sides of the tablet.

During the beginning of the ejection, it can be seen that the surface of the cup is moving whereas the land is not. This represents the deformation of the tablet. The natural consequence of this deformation is an increase of the shear stress on the upper side of the tablet. At the same time, the reverse trend is observed on the lower side of the tablet, i.e. the shear stress decreases.

At some point, the force needed to move the tablet is reached, and the tablet begin to move upward. During this movement, the value of the shear stress remains nearly constant on both sides of the tablet.

This FEM simulation of the first part of the ejection process makes it thus possible to understand the influence of the ejection on capping of biconvex tablets. During the first part of the unloading, the shear stress, that is responsible of the failure of the compact, increases on the upper side of the tablet and decreases on its lower side. The shear strength of the tablet may thus be reached because of the deformation of the tablet. This would promote capping only on the upper side of the tablet as sometimes observed during manufacturing.

This mechanism makes it also possible to understand the influence of various process/product parameters on the capping tendency. As explained above, the deformation of the tablet during the beginning of the ejection explains that capping occurs on the upper side of the tablet. The deformation of the tablet is due the force applied by the lower punch. It is obvious that a higher force applied by the lower punch promotes a greater deformation of the tablet. So, all product/process parameter that would increase the ejection force would favor capping.

First, ejection force obviously increases along with the friction coefficient between the tablet and the die. A high friction coefficient is likely to promote capping. This result indicates that a bad lubrication, which will increase the friction coefficient, increases the capping tendency of biconvex tablets. This trends is already well known and the mechanism presented in this presentation explains it.

The second point, which is of great importance, is the influence of speed. It is well known that increasing the speed favors capping occurrence. In fact, increasing the compression speed increases the ejection force. Some authors explained the increase of the ejection force along with the speed through problems of lubricant migration during the compression. If this explanation can be imagined there is, to our knowledge, no experimental data to prove this hypothesis. Another explanation for this phenomenon can be proposed. This increase could be simply due to an inertia effect. In fact due to the high speed of the punch, there is an overshoot due to the fact that the force is not transmitted instantaneously. This fact can be observed experimentally.

When slow speed is used, the ejection signal is similar to the one obtained in the simulation. The force increases and then reaches a plateau as predicted by the coulomb friction theory. On the contrary, the ejection signal obtained at high speed has a totally different shape. First of all, the peak force is much higher than in the case of a low speed. Moreover, after the peak, the force oscillates. This pattern is typical of a stick-slip phenomenon. This signal means that during ejection, the tablet is subjected to an oscillating deformation that will induce successive increase and decrease of the shear stress on the top of the tablet. This kind of mechanical solicitation is of course not in favor of the tablet integrity and is likely to increase the capping tendency.