Process Control – A Closer Look

To effectively control a process, a deep understanding of the process is a must

Why is Process Control So Important in Pharma Manufacturing?

Process control forms the backbone of any manufacturing process to ensure quality and consistency of the products produced. In pharma, the stakes are high as the products are intended to treat disease in humans. And so, it is essential to ensure the process stays within predefined control limits to guarantee the safety and efficacy of the final therapeutic product, and to ensure regulatory requirements are adhered to. A process that is poorly controlled can lead to out-of-spec product, regulatory issues, product recalls, not to mention result in significant waste of valuable resources and subsequent market delays, which in-turn fuels the ever-growing global issue of drug shortages!

Process control is associated with all stages of the drug development lifecycle, from drug substance development right though to drug product manufacturing and forms a key part of the control strategy, which is defined and submitted as part of a regulatory submission.

How is Process Control Implemented?

To effectively control a process, a deep understanding of the process is a must. The core of this knowledge is established during early process development studies, where critical process parameters (CPPs), the variables that directly affect the critical quality attributes (CQAs) of the product are identified, as well as other key functional relationships between key process and material inputs and outputs. Analytical tools and measuring devices, whether they are simple sensors measuring temperature and humidity, or more advanced Process analytical technologies (PAT) for example: the Eyecon2 particle size and shape analyser developed by InnoGlobal Technology (formally known as Innopharma Technology) or spectroscopic technologies such as Ramen, play a crucial role in capturing key material and process data in addition to the data generated by the manufacturing equipment. Together, these tools provide powerful, real-time process insights that accelerate the process learning curve.

?Traditional process control relies on carefully following a process recipe, where the operator observes the process and manually adjusts equipment parameters to maintain the process within defined control limits. This of course has its drawbacks, as it’s pretty subjective and may vary from one operator to the next, affecting process reproducibility.

As the Pharma industry continues to adopt more advanced process control and automation methods associated with Smart Manufacturing and the more recent concept of Smart Factory, equipment is integrated / interconnected with IIoT (industrial internet of things) sensors and other advanced technologies to enable real-time process monitoring and control. These advanced control systems are carefully configured to automatically respond to any material, equipment or environmental disturbances, by making automatic parameter adjustments to ensure the process stays within the quality target product profile (QTPP).

SmartX is a cloud based automation and digitalisation platform developed by InnoGlobal Technology, which facilitates the integration of PAT and IIoT sensors with process equipment, enabling real-time process control and automation. A low-code / no-code control module enables the implementation of? advanced process control without the need for programming skills. Let’s now take a look at a couple of process control examples that were implemented in a fluid bed Wurster coating process through the low-code / no-code control module of SmartX, but first, a quick overview of Wurster coating.

Wurster Coating Process Overview

Wurster Coating describes the layering of mulitparticulate cores with a drug or a controlled released coating for example, in a fluid bed Wurster coater. The equipment consists of a product container with an air distributor plate at its base which supports the batch of raw multiparticulates. Heated air passes up through this bottom plate and fluidizes the bed of multiparticulates. A spray nozzle is located at the center of the air distributor plate, with an upward spray direction. Directly above this is the Wurster column, fixed at a certain distance above the bottom plate to allow material to pass up through the spray zone in the center of the Wurster column. Here the particles are coated with tiny, atomized droplets of coating formulation. As the particles exit the top of the Wurster column they are dried in the heated air stream, decelerate, and fall down the container sides to rejoin the particle down bed. This cycle is repeated many hundreds if not thousands of times over several hours, until a predefined amount of coating is applied, known as percentage weight gain. A filter situated above the product container allows moist air to escape while retaining the multiparticulates within the process.

Some CPPs of Wurster coating include: the rate and temperature of the air introduced through the bottom of the container, the spray rate of the coating formulation, the atomising pressure applied to the spray nozzle (forms the tiny spray droplets). These parameters are interrelated with complex relationships and so cannot be thought of in isolation, instead they require careful fine-tuning and optimisation in relation to one another. An example of a Wurster coating CQA is dissolution i.e. the release rate of the drug, which is directly related to the thickness of the coating layer applied to the multiparticulates.

Example 1 - Controlling Spray Rate

Optimum spray rate is important to ensure sufficient coating formulation is delivered to the surface of the particles on each cycle through the spray zone. Too low a spray rate can result in spray drying, where the coating solution dries in the heated air stream before reaching the particles surface. If too high, then over-wetting can occur leading to agglomeration of the particles known as twinning. The fluid bed equipment used in this example i.e. a GPCG2 lab scale system, reports spray pump speed (%), as opposed to spray rate (g/min) as used in real world processes. The optimum spray rate for the process in this example is 12g/min. Manual control of the spray rate would first require calibrating pump speed against the total number of grams of coating solution sprayed over e.g. a 1 minute period, as measured from the loss-in-weight of the coating solution sitting on the scales. Periodic verification of the loss-in-weight per minute, against the pump speed setpoint is required during the process to avoid drift, with the operator adjusting pump speed as needed to ensure a spray rate of 12g/min is maintained.

In the example outlined here, the scales and fluid bed equipment are integrated within the SmartX platform, enabling the use of a specific control function developed within the low-code / no-code control module, to automatically set the pump speed needed to achieve a desired spray rate. The control function requires the user to specify certain input values, such as the desired spray rate (g/min), as well as other equipment and formulation related values. Based on these inputs and the loss-in-weight data acquired from the scales, the control function predicts and implements the required pump speed needed to achieve the target spray rate. An integral error component built into the control function works to continually adjust pump speed to account for any error or deviations in the setpoint, to ensure the target spray rate is maintained. This is referred to as a closed-loop control system due to its self-regulation. As a result, process robustness, reproducibility and mitigation against potential human error is achieved, helping to ensure product quality is maintained.

Example 2 - Controlling Airflow Rate Based on Particle Size

The second example focuses on controlling airflow rate (m3/h), which should be optimized to maintain particle fluidization during the process. Too little airflow and the bed of multiparticulates will not fluidize sufficiently, while too much airflow will force the particles up into the exhaust filter affecting final yield and may also contribute to spray drying, compromising coating formation. In addition, as the coating cycle continues, the particles increase in size on each pass through the spray zone. To counter the effects of the increasing weight of the particles, airflow must be adjusted during the process to maintain sufficient fluidization. In a low weight gain, single stage process like in this example, the increase in particle size and weight is relatively small. However, many industrial processes can undergo several independent coating steps, each with different weight gain targets lasting hours to achieve the desired final product, so the effects become more apparent. Regardless of scale the requirement is the same.

In a manually controlled process the airflow is adjusted by the trained eye of the operator, by periodically observing the flow pattern / height of particle fluidization through a process window and adjusting airflow rate accordingly within predefined limits. The example here uses an advanced control approach, where optimum airflow is controlled based on real-time particle size (Dv50), as measured by the Eyecon2 particle size and shape analyser. Dv50 refers to the mass median particle diameter. As before, the control function was implemented through the low-code / no-code control module within SmartX. The control function calculates the velocity of air needed to optimally fluidize particles of a given size, as measured by the Eyecon2. The airflow rate required is then determined based on this velocity and the cross-sectional area of the product container. This airflow rate is relayed as the target setpoint to the fluid bed system for implementation. This approach enables the control system to automatically set the optimal airflow rate for a specific size of starting material regardless of size variations, and also accounts for increasing particle weight during the process. This approach can be just as easily applied to larger scale equipment, as it factors in the change in the cross-sectional area which scales with increasing size.

Conclusion

Robust control is an integral part of pharmaceutical manufacturing and requires a thorough understanding of the process and in-process materials, which as you can see from the above examples can be quite complex. The use of advanced process control methods can ensure the process is continually monitored and adjusted by the control system in response to real-time deviations / disturbances, ensuring the quality target profile is maintained and the quality of the end product is guaranteed. This helps to address quality issues and prevent drug shortages, in addition to reducing waste of valuable resources associated with failed and repeated batches. Systems like InnoGlobal Technologies SmartX Automation and Digitalisation Platform can help support this journey through seamless integration of IIoT sensors with manufacturing equipment, and by facilitating both the configuration and implementation of advanced process control.

About the Author:

Caroline is a Process Scientist at InnoGlobal Technology with 7+ years’ experience, with sometime in medical device manufacturing. Her academic background includes Biochemistry, Pharmaceutical and Medical Device Manufacturing, and Industrial Biopharmaceutical Analysis. Caroline is responsible for experimentation around our process analytical technologies and advanced process control techniques currently under development, across the oral solid dose manufacturing space.

Contact us today to learn how InnoGlobal Technology can transform your manufacturing operations. [email protected]