Total laboratory automation in the pharmaceutical industry

It is abundantly clear that, across a range of industries, the uptake of automation has been immense. But what does that mean for the modern manufacturing chemist? Joseph Willmot, Applications Leader, and Dr Paul Orange, Chief Marketing Officer, H.E.L Ltd, investigate

Automation is, by definition, the use or introduction of automatic equipment in a manufacturing or other process or facility.

To the contemporary chemist, this means less hands-on time with a specific procedure, which, in turn, can lead to higher productivity or a more varied workflow.

Historically, chemical reactions have always been tricky processes to automate. The number of dynamic variables that are continuously fluctuating throughout the duration of a reaction predicates the need for complex and comprehensive control systems.

Only in recent years, with the rapid development of computing power and software control, has this been viable. Some of the variables that require control can be seemingly simple parameters — such as temperature or pressure.

However, these factors may be profoundly significant to the yield or safety of a process, for example, by inhibiting side reactions or preventing the thermal decomposition of a product. Because of this, each chemical reaction needs thorough characterisation before scale-up can even be considered.

But how does a chemist, with only one set of hands, comprehensively characterise reactions?

The answer is automation. Using this powerful tool, the chemist can run multiple experiments under different conditions to fully understand the dynamics of the reaction.

Design of experiment (DoE) techniques utlise automation alongside sophisticated modelling and applied statistics, which allows for informed decisions to be made regarding which chemical reactions can be progressed to the next stage of process development.1

Total laboratory automation (TLA) is now possible for individual clinical laboratories.2 Using TLA, all input processes such as sample preparation and barcoding, right through to the output processes such as report generating and data analysis, are all automated.

This level of automation has shown significant improvement in turnaround times for samples, leading to cost savings and the increased accuracy of results.3

Manufacturing automation

The opportunities with automation only grow with scale: trying to cool one litre of liquid is a completely different problem compared with trying to cool 5000 L.

Because of this, intense process development is used to fully understand the chemical process at the laboratory scale. Only then can scale-up occur.

However, scale-up itself can present complexities. Generally, physical properties are slower to respond, even if appropriate scaling has been applied. An example of this is temperature control. A 1 L reactor that is 80% full with 60 °C water will cool to ambient temperature at a rate of approximately 0.5 °C/min.

But a 5000 L storage tank under the same conditions has a cooling rate of 0.023 °C/min (both rates are dependent on the insulation of the vessel). This significant difference highlights the drastic modification needed to the control system.

The 1 L scale exothermic reaction would require a relatively small amount of cooling to keep the process isothermal. Normally, with scale-up, there is a decrease in the efficiency of heat transfer. Because of this, the final cooling duty required will be greater than the volumetric scale-up. There is also the need to include an appropriate safety factor, futher increasing the cooling duty required.

Laboratory tools such as adiabatic calorimeters can be used to determine the cooling capacity required to safely run manufacturing processes.4 These have the additional benefit of being able to model and apply worst-case scenarios directly at the manufacturing scale, preventing the need for manipulating and scaling data from laboratory scale processes.

Safety with automation

Any process must be designed with safety as a key priority. The controlling system and automation of the process must, therefore, also reflect this. Automatic shutdown procedures provide a desirable level of protection. These can be user-defined and based on the current operating conditions of the system.

For example, the feed for a reaction can be programmed to turn off automatically if the system temperature was to rise above a defined limit, thus allowing the temperature control system to regain safe operational conditions.

Additional levels of protection can be set, so that if the first shutdown procedure was insufficient and the temperature continued to rise, emergency cooling would activate. If the temperature raised even further, a quenching agent could be activated.

In the most simple scenario, this automation can come from physical fail-safe devices such as pressure relief valves (PRVs), which need to be appropriately sized and modelled from vent sizing calculations based on laboratory scale experiments.

PRVs open automatically during a dangerous pressure rise and are engineered to ensure that the pressure rating of the containment vessel is never exceeded.

System failure is an additional hazard risk and should be a critical consideration in automation.

If an individual component loses software communication or signal with the controlling system, it needs to work as an independent safety system.

This has to be included in the original design. For example, should a control valve be a fail closed or fail open one? If it is positioned on a feed pipe, a fail closed state makes sense. If on an automatic vent, then a fail open state might be more appropriate.

Another consideration is ensuring that the rest of the controlling system takes appropriate action based on a loss of communication. With the increasing power of software, automatic “WatchDog” functionality is present in the majority of equipment. In this, a set point is preloaded to the device and utilised if communication to the controlling software is lost.

Summary

With these automatic safety procedures in place, not only can the manufacturing chemist run a process with fewer manual interactions, they can also leave the system unattended, knowing there are appropriate protocols in place to protect the system in a worst-case scenario.

As the power of automation increases, so will the capability of the modern chemist. At the laboratory scale, more samples can be tested with better accuracy and granularity for an increased understanding of the chemical process.

During scale-up, automation allows for safe design of the manufacturing procedure, identifying possible hazards and risks. At the manufacturing scale, automation allows for safe processes to be completed day in, day out, without any loss of productivity, quality or system control.

References

  1. V. Rosso, et al., "Uniting Laboratory Automation, DoE Data and Modeling Techniques to Accelerate Chemical Process Development," React. Chem. Eng. 4, 1646–1657 (2019).
  2. J.R. Genzen, et al., "Challenges and Opportunities in Implementing Total Laboratory Automation," Clinical Chemistry 64(2), 259–264 (2018).
  3. M.L. Yarbrough et al., "Impact of Total Laboratory Automation on Workflow and Specimen Processing Time for Culture of Urine Specimens," Eur. J. Clin. Microbiol. Infect. Dis. 37, 2405–2411 (2018).
  4. www.fauske.com/blog/safer-scale-up-of-batch-and-semi-batch-reactions-part-2-quantification-of-the-desired-reaction.

Companies