Medical Imaging Monitors Pharmaceutical Process Health
Tomography is widely used in medical imaging, but also is becoming a useful PAT tool for process development and optimization. In this article, Ken Primrose, principal of Industrial Tomography Systems, discusses electrical resistance tomography (ERT) and its implications for pharmaceutical manufacturing.
By Ken Primrose, Industrial Tomography Systems
It’s relatively easy to get on-line data from a point source in a process vessel. But the data are of little value unless your liquor is homogeneous, so that the sensor sees a representative sample of the contents. And how do you know if you have perfect mixing? Scanning the whole volume would be much more valuable, and that’s just what a new technology – electrical process tomography (ERT) – does.
The FDA’s encouragement of science-based approaches to production control (under its Process Analytical Technologies initiative) is a great stimulus for such innovative ideas. Where process conditions in a vessel vary, using ERT to get spatial rather than point-based information can make a big impact on process control and optimisation.
|Figure 1: How ERT visualizes material distribution in a process vessel.|
ERT is the process industry equivalent of whole body scanning in a medical environment. It is non-intrusive and has no moving parts. Industrial Tomography Systems (ITS) of Manchester, UK – a spin-off from the University of Manchester – is developing the technology as a PAT tool. It has on-going projects under confidentiality agreements with a range of international companies, amongst them some giants of the pharmaceuticals industry.How the technology works
What does ERT look like in action? It is fundamentally a simple sensor network coupled with highly sophisticated electronics and software. Imagine an array of electrodes — typically 16 — arranged in the same plane on the inner circumference of a process vessel. If you then apply a small AC current very rapidly between a pair of electrodes, the resultant voltage difference between remaining electrode pairs can be measured according to a pre-defined measurement protocol. Then apply current through the next pair of electrodes and repeat the voltage measurements. And so on.
An image reconstruction algorithm, the heart of ERT, provides amongst other things an image of the distribution of materials in the sensing volume (see Figure 1, above). Add other sensor arrays and you can view the situation throughout the whole vessel.
How many measurements and how fast? In the course of some 25 milliseconds a 16-electrode sensor gives 104 independent measurements – enough to get a real-time picture of the conductivity of the process fluid in that volume. Conductivity can be related to many physical properties, and in particular to the concentration of different components.
|Figure 2: ERT electrodes arranged on a probe.|
Placing electrodes around the internal surface of a vessel may be a rather laborious use of resources. So ITS worked on an alternative. Inserting electrodes along the length of a probe (Figure 2, at right) creates a much more versatile piece of equipment which can be lowered into a vessel in place of an existing baffle. A probe can be used across a range of vessel sizes. Of course the image reconstruction algorithm needs appropriate adjustment for the new disposition of electrodes, but the result is a similar real time image of the contents of the whole vessel.
In various cases, electrode deployment matches other practicalities of the situation. To picture fluid flow in a pipe, the most appropriate array is similar to that around the internal circumference of a vessel. In a hydrocyclone the inside wall is again the position of choice, but with a graduated diameter. In pressure filtration, which is described later in this article, arranging them across the filter bed proves least obtrusive.And how would you use it?
All sorts of applications spring to mind. Throughout the pharmaceutical industry there is a need, for instance, to ensure good dispersion of solids in liquids or to observe separation of organic and aqueous phases. Current technology is limited in what it can tell you. ERT allows you to watch the mixing characteristics of the process vessel contents while you adjust conditions to improve them (Figure 3, below).
Crystallization is another process well suited to monitoring by ERT. When two ionic species in solution produce a salt, the difference in signal between salt and solution is rather large. This means that ERT is very sensitive to the onset of crystallization. Analysing the dissolution of slow-release capsules is another potential application. These capsules are meant to let drugs out slowly into the body from the stomach and have been in the news recently (www.pharmacist.com/articles/h_ts_0747.cfm
). During proving tests in a “simulated stomach,” ERT could provide a clear picture of how salts are released from a tablet.
As a substitute for tradition
|Figure 3: The effect of different mixing speeds revealed by tomography.|
Pressure filtration is an acute example of a process needing quality on-line monitoring. At the moment, there is little or no on-line help in determining end point or assessing filter cake homogeneity. Traditional methods are basically empirical. Filtration routinely continues far too long, wasting time. And there is no satisfactory way to spot formation of rat holes in the filter cake.