Understanding the continuous process and the challenges it poses

In this article, Jamie Clayton, operations director, Freeman Technology, considers the challenge of implementing continuous tablet production and evaluates dynamic powder characterisation as a technique to support the development of a continuous wet granulation process.

Rising development costs, the requirement for faster times to market and the growth of generics production, have contributed to the need to improve efficiency across the pharmaceutical life cycle. Quality by design (QbD) and the application of process analytical technology (PAT) are driving progress, and smarter approaches to process development and operation, such as continuous manufacturing, are gaining momentum.

Batch production still dominates as it is well established and has recognised benefits. Manufacturing discrete batches makes it easier to isolate problems, and stringent regulations can also deter development of new practices. However, even a well-optimised batch process can suffer inefficiencies; batch variability and products out of specification are not uncommon.

Continuous processes provide benefits such as reduced labour requirements, lower capital investment and simplified scale-up at the same time as addressing problems associated with batch variability to ensure consistent output. However, developing continuous processes that deliver these benefits is a challenge that requires a comprehensive understanding of both the process and the materials.

Many pharmaceutical manufacturing operations can already be considered semi-continuous; roller compaction, milling and tabletting, for example. However, the finished tablets are the product of several sequential operations which together combine active pharmaceutical ingredients (APIs) and excipients to deliver a tablet with the required critical quality attributes. Figure 1 illustrates the typical stages in tablet production, which in batch manufacturing are managed on a discrete basis.

The first step is often wet or dry granulation, to convert often dissimilar excipient and API particles into a homogeneous, granulated mass to avoid downstream segregation and improve processability. In wet granulation, the exiting mass is dried to remove excess moisture, and in both cases the granulate is typically milled prior to the addition of further additives to enhance performance in the tablet press. To ensure efficient processing, the output of each of these operations must be optimised for the next stage and in batch manufacturing, the approach is to apply well-defined procedures, assess the outcome, apply any remedial action, then move material to the next stage. This requires manual intervention and off-line analysis.

Continuous manufacturing involves integrating each of the process steps, making it essential to control each in real time to ensure a suitable output for the following stage, with minimal intervention between stages. However, the various processes employed in tabletting will subject powders to a range of stresses and flow regimes, such as forced flow through an extruder, unconstrained flow into an empty tablet die, and compression. A successful process therefore requires a comprehensive understanding of how material properties and process parameters influence the output of each stage.

In wet granulation, for example, screw speed, feed rate and water content all contribute to granulate properties so success relies on understanding how a powder responds to these variables and how to manipulate them to produce granules that are optimised for subsequent steps.

It’s increasingly acknowledged that traditional powder tests, such as angle of repose, flow through an orifice, and tapped density measurements, provide limited information. They may provide some insight into powder behaviour but are often compromised by poor repeatability and reproducibility, as well as a lack of sensitivity. Furthermore, the test conditions don’t represent the conditions in process, limiting the ability of these methods to deliver relevant information.

In contrast, dynamic flow testing measures various properties of a powder in motion, and under conditions that simulate the process environment, providing relevant information that rationalises in-process behaviour.

Basic flowability energy (BFE) is determined by measuring the axial and rotational forces acting on a rotating blade as it descends through a powder (Figure 2) and reflects how a powder will flow under forced conditions, such as those in an extruder or feed frame. Measuring BFE in different states quantifies the impact of processing conditions such as consolidation and fluidisation.

Case study: Applying dynamic powder testing to define a wet granulation design space

Studies were undertaken to quantify the quality of granules produced during wet granulation. The aim was to investigate relationships between dynamic flow properties of granules and critical quality attributes of the resulting tablets.

A GEA Pharma Systems ConsiGma 1 high shear wet granulation and drying system was used to produce granules and BFE was measured at four stages: on exit from the granulator; post-drying; post-milling; and following blending with a lubricant. Two different blends were evaluated: one based on paracetamol (APAP) the other on dicalcium phosphate (DCP).

The first investigation assessed the impact of varying water content, powder feed rate and screw speed on the BFE of the resulting granules. Figure 3 shows how BFE of the APAP granules produced at different screw speeds varies as a function of water content.

At constant screw speed, increasing water content, results in a higher BFE. Decreasing screw speed, at a given water content, produces granules with a lower BFE. This is consistent with the observation that higher water content and lower screw speeds tend to produce larger, denser and more cohesive granules that present substantial resistance to blade movements. Furthermore, granules produced at a water content of 11% and screw speed of 600 rpm have similar BFE values to those generated using a screw speed of 450 rpm and a water content of 8%. This introduces the possibility of producing granules with similar properties using different process conditions, an important factor in defining the operating scope.

Figure 4 shows the impact of dry powder feed rate. The BFE of DCP granules, produced with a fixed water content of 15% and constant screw speed of 600 rpm, was found to be indirectly proportional to feed rate. Reinforcing previous findings, granules produced at a feed rate of 18 kg/hr with 15% water had similar properties to granules containing 25% water made at a feed rate of 25 kg/hr.

Table 1 shows BFE measurements for two pairs of granules, each pair having similar BFE values but produced using different conditions. Conditions 1 and 2 result in granules with a BFE ≈ 2,200 mJ, while 3 and 4 have a BFE ≈ 3,200. The BFE of these granules was also measured at the four stages detailed above.

Figure 5 shows how the BFE changes following each stage of the process. Granules made using conditions 3 and 4 show an increase in BFE after drying while those produced using conditions 1 and 2 remain relatively consistent. This increase can be attributed to the granule’s relatively large size which, in combination with increased density and hardness, results in increased mechanical interlocking when dried, and greater resistance to forced flow. The granules produced using conditions 1 and 2 have a weaker structure, lower density and relatively small size, and are therefore less prone to mechanical interlocking. The BFE of these samples therefore only changes slightly following drying. After milling, particle sizes are more similar and flowability for the four batches converge, although differences in granule density, shape and stiffness still exist, rationalising the differences in BFE which are retained following lubrication.

These results demonstrate the flexibility to produce granules with specific flow properties. The next stage was to determine whether such properties could be targeted to ensure tablet quality.

Tablets were produced from the four sets of lubricated granulates and Figure 6 shows the strong correlations between the BFE of the granules at each stage and the hardness of the resulting tablets. This demonstrates how operators can ensure a critical quality attribute of the finished product by adjusting critical process parameters to target properties of an intermediate. This has the potential to accelerate development and scale-up and facilitate efficient process control during routine operation.

The need for increased efficiency throughout the pharmaceutical lifecycle is encouraging the industry to innovate new manufacturing practices. The economic gains of continuous manufacturing are a considerable attraction but the challenge of developing successful processes is not inconsiderable. This case study highlights the relationship between the flow properties of a formulation and the critical quality attributes of the resulting tablets, demonstrating the potential role of dynamic powder characterisation in advancing manufacturing processes and the implementation of continuous manufacturing.