Fluid bed drying

Process Analytical Technology (PAT), the final frontier. These are the voyages of FDA’s initiative to provide a framework for innovative pharmaceutical development, manufacturing and quality assurance … and boldly go where no regulatory agency has gone before.

Before this programme was launched, more than 10 years ago, pharmaceutical production relied on process validation and end product testing, complemented by in-process controls (IPCs). The idea was that if every process parameter was maintained within a given, validated range, the intermediates and final products would always comply with the required quality specifications. However, this approach relied on a validated array for each and every parameter that could potentially influence the operation.

If we look at a real-life example, such as dispensing, predetermined quantities of the ingredients in a specific formulation are weighed into an intermediate bulk container (IBC). Assuming a total amount of 300kg, a further 45kg of binder solution is added when the mixture is processed in a high shear granulator and subsequently transferred to a fluid bed dryer. Here, the material needs to be dried to obtain a moisture content of, for instance, <2.5%.

As the drying process is itself validated, the material will be exposed to 4,500m³/h of drying gas at an inlet temperature of 60°C for 50 minutes. Surprisingly, the material is not simply discharged after that time but, in nine out of ten examples, the process is merely paused while an operator removes a sample for offline moisture testing. If the material happens to be sufficiently dry, an extra 10 minutes of time may elapse; but if the material is still too wet, drying continues for a few minutes more before another sample is taken … and another 10 minutes pass.

If the specification simply demands a moisture content of less than a certain percentage (2.5% in our example), this procedure works – but productivity suffers. The fluid bed dryer is just acting as a simple holding container during what is essentially downtime. If, however, the target moisture content requirement is more complex and must fall within a specific range – 1.5-2.5%, for example – the situation becomes much more challenging.

Chemical stability
And, last but not least, a moisture content of less than 2.5% ensures chemical stability, which results in an acceptable product shelf-life. However, the next unit operation in the process train is tableting, which in the majority of cases demands a granule-to-humidity ratio at the upper end of the range to minimise problems such as capping.

So where does the variability come from in a validated process such as this? It shouldn’t happen. The simple answer is that all potential sources of variability are rarely covered by the validation. Raw materials are often stored in non-classified areas such as warehouses. For instance, when starch – which has a high moisture equilibrium – is added to a process, the actual amount of solid material may vary according to the humidity at which it was stored: a variable percentage will be water.

In many cases, the inlet gas is not humidity controlled, which means its drying ability will depend on the external environmental conditions (summer/winter effect). A third source of variability is tablet production, which is often done in campaigns. Batch number one may be done in completely clean equipment, resulting in a yield of approximately 95% (a lot of material sticks to the surfaces), whereas the following batches may have yields of 100%.

FDA was well aware of these types of issues when it launched the PAT initiative. With the objective of modernising pharmaceutical production, manufacturers and agencies were asked to focus on critical quality attributes (CQAs). In fluid bed drying, this is clearly the final moisture content, as well as (albeit to a lesser degree) parameters such as gas flow rate, inlet gas temperature and drying time, which were key elements of the classical validation approach.

Importance of measuring critical parameters
If the critical parameter can be measured directly, the other parameters can play a smaller role. With end-of-drying product moisture content, this means that we need a method to determine the value directly. As proven many times in the food or chemical sectors, this can be done using NIR spectroscopy. Plus, as we only want to monitor one value, there’s no need for complex chemometrics. A simple calibration over the relevant moisture values is enough. With the Lighthouse Probe, GEA offers a proven, off-the-shelf solution to avoid window fouling. When combined with a basic NIR device, moisture levels can be determined in real-time, drying can be terminated when the desired moisture level has been reached and offline testing can be relegated to ancient history. Welcome to the 21st century.

Dr Harald Stahl is with GEA

 

Recent Issues