Optimizing Tablet Compression

In an actual pharmaceutical manufacturing environment, however, the number of stations plays a somewhat less important role regarding the number of tablets that can be produced with a particular press. The majority of pharmaceutical tablets are not produced at the press’s maximum compression speed because it is not possible to produce tablets of acceptable quality at high rotation speeds. Defects such as capping, sticking and lamination occur, and the tablets become subject to weight and content variations. In many cases, reducing the rotation speed of the press makes it possible to avoid these problems. As such, the simple relationship between reduced rotation speed and fewer out-of-specification tablets will be observed.

Usually, eccentric presses or small, rotary tablet presses operated at slow speeds are used when developing tablets. Often, the development process focuses on optimizing tablet characteristics such as hardness, disintegration time, stability and/or friability. Less attention is paid to the tableting process itself, as very different operating parameters will be used during full-scale production. Typically, tableting problems only occur during scale-up and when using high-speed production-scale machines. This phenomenon will be explained below using tablet capping as an example.

If a given substance is exposed to a pressure load, it will react in a number of different ways

• if a modelling material is deformed by mechanical energy, for example, the mass will maintain its form even when the external force is no longer applied; this is called plastic deformability
• if a spring is deformed mechanically, it will return to its initial state even when the external force is no longer applied; this is referred to as elastic deformability
• if a mechanical load is applied to cornflakes, for example, the phenomenon of brittle fracture occurs
• in addition, viscoelasticity — a combination of the previously mentioned reactions — can occur; this describes substances that react either plastically or elastically depending in a time-dependent manner. An illustrative example of this is inflating a bicycle tyre with a hand pump: if the load is applied slowly, the piston can be pressed down and the air will pass into the tyre; if the piston is depressed too quickly, the air will not flow into the tyre and the pump reacts elastically.

The extent to which a tablet is prone to capping depends on the deformation behaviour of individual components. If materials are used that deform plastically or undergo brittle fracture, the risk is low. But, if the tablet formulation contains substances that deform elastically or demonstrate viscoelastic deformation, there is a high risk of capping, particularly with rapidly applied loads. The situation is exacerbated if the active pharmaceutical ingredient (API) itself shows this behaviour and has to be incorporated into the tablet in high concentrations. In almost all other cases, capping can be avoided completely by the appropriate choice of pharmaceutical excipients. However, capping will always occur if, following compression, more elastic energy is accumulated in the tablet than its inner structure can absorb.

Apart form the choice of excipients, the processes that precede tableting also influence the tablet’s tendency to cap. In the case of direct compression, only the compression properties of the substances used will define the extent of the capping potential. Another problem associated with direct compression is the higher proportion of fines, which also increase the tendency to cap. Wet granulation, by contrast, enables capping to be minimized, according to how evenly the binder is distributed during granulation. Therefore, granulates that have been produced by spray granulation generally cap less than those produced using an intensive mixer granulator.

Another cause of capping is entrapped air that has been compressed during main compression and eventually shatters the tablet as a result of perfect elastic behaviour. The more open-pored a material is — usually discernible by its low bulk density — the more air it contains. The majority of this air should be removed during pre-compression. Yet, the problem here is that with faster tableting speeds, less time is available. Different tablet press manufacturers have developed a variety of concepts to improve this situation; as such, the speed of the press can be increased by up to four times for critical formulations [1].

If the diameter of the tablet press rotor is X cm, then the die covers a distance of S = X * π cm during one rotation. The circumferential speed (V), measured in m/s, can be calculated as follows: V = S * rpm/60 (with rpm being the number of rotations per minute). The division by 60 is necessary to define the speed in m/s.

If a press is operated at a lower rotation speed to avoid problems with capping, this can be equated to a reduction of the circumferential speed. In other words, capping can be prevented if the press runs below a certain circumferential speed. Thus, if a formulation has a strong tendency to cap, the number of tablets produced per hour cannot be improved by simply increasing the number of pressing stations. Only a reduction in the distance between the dies, which is offered by several manufacturers, can improve the output (Table I) [2,3].

Press A is the reference. Press B is identical, but has a bigger rotor. Press C has the same rotor as press B, but the number of press stations has been increased by reducing the distance between the single dies [2,3]. Assuming that the maximum circumferential speed is 2.5 m/s owing to capping, this results in the fact that the larger presses (B and C) have to operate at a reduced rotation speed compared with Press A. As a result of linear correlation, in the case of press B, this counteracts the effect of the increased number of pressing stations. The higher output achieved with Press C is the result of the reduced distance between the single dies.

Usually, it is not possible to change the formulation during scale-up from R&D to production to reduce the tendency to cap. And, in most cases, only minor adjustments can be made to optimize upstream processes. Apart from reducing the number of turret rotations, and therefore the circumferential speed, only two other options remain. On one hand, punches with larger heads can be used. And, on the other hand, it is possible to retract the upper punch more slowly following main compression. As a result, in many cases, the stored energy can be transferred to the upper punch without the tablets being destroyed by capping.

Each tableting process aims to produce tablets with a constant weight. Yet, as a result of variations in the density of the feed material and partial or incomplete filling of the dies, there are always weight variations (the relevant pharmacopoeia specify acceptable weight variation levels). The threat of weight variation is minimized if the feed material is produced by granulating or compaction; ideally, the composition of the feed material should be determined down to single particle characteristics. But, if the feed material has a wide particle size and/or density distribution, the risk of segregation and subsequent weight — and tablet content — variation is high. This danger can be minimized by mechanically decoupling the press and the feed material to minimize the risk of segregation. Furthermore, allowing the feed material to free fall between unit operations should be avoided.

Similar to capping, content variations are more pronounced at higher press speeds: with increasing rotation rates, the ruling speed also increases, which means that the period that the die remains under the filling unit decreases. This means that with increasing press speeds, greater demands must be made on the flowability of the feed material. An alternative approach would be to impose a maximum circumferential speed for each powder flowability rate to guarantee uniform die filling.

There are different ways to characterize flowability, including the Hausner factor or by determining the angle of repose, and one of the key tasks of developing mainstream processes must be to significantly improve the flowability of the feed material. A detailed consideration would go beyond the scope of this article; but, generally, every effort should be made to granulate the material with the greatest possible mechanical energy while minimizing the formation of lumps and caking. This material must be milled again during downstream processing, which results in an increased amount of fines and poor flowability.

During scale-up from R&D to production, upstream processes can only normally be optimized within very narrow limits. But, often, the situation can be improved by using a ‘forced filling’ approach. If the lower punch is retracted before the die reaches the filling unit area, the material enters the die as a result of gravity. With forced filling, however, the lower punch is flush with the die table. The lower punch is then pulled into its target position below the filling unit. The material is sucked into the die because of the resulting vacuum, which makes it possible to use high pressing speeds even if the material doesn’t flow optimally.