How Filters Work: Mechanisms of Filtration Part 1:

In the last blog, we recognized there are a great number of filtration applications, each slightly different because the job they need to perform is unique to the application, and even the exact process they are installed within.  Despite this variety, they all share common foundations, and they all rely upon a small number of basic principles to achieve their desired performance.

Direct interception: The first mechanism of filtration

When we visualize a filter, we often think of something like a kitchen sieve or colander.  This simple visualization highlights the two main components of a filter, the material it is made from and the holes, or pores, within it. And it is the size of these pores that leads to the first mechanism of filtration.

Big, non-deformable particles cannot fit through small holes.  We know this intuitively and we see this first-hand when we use a colander to retain cooked pasta from a pan of boiling water. But even this simple physical truth has additional layers of complexity. Using the pasta analogy, something like a penne pasta will be reliably retained – we can call this absolute retention.  If we use the same colander for spaghetti, it is not impossible to believe that a single spaghetto (yes, that is the name for one spaghetti!) could be washed through one of the holes, or that smaller pasta, like orzo, would pass through relatively easily.  Direct interception (size-based exclusion) is therefore most often dictated by size, but the particle shape may also play a part in whether it is retained.

Interaction between the particle and the material of the filter can also be factor in retention. Imagine orzo again, but now only a single very overcooked, sticky particle!  In a fast-moving stream of water, this may still flush through the pores but without enough water (or none) it may stick to the colander and be retained even if it is a long way from the pore.  This type of interaction is very real at the microscopic level.

Finally, we look again at our orzo process.  A single particle may pass though the pores of the colander, but with a large quantity of orzo being thrown into the same colander at the same time, the orzo particles interact and may bridge the pores and be retained while the cooking water passes through.  We will observe the majority of the pasta being retained but a small number still passing through the pores, probably early in the process. This interaction tells us that filter performance can be influenced by the actual process conditions as well as the size of the particles and the size of the pores.  Of course, in the case of our precious orzo, we could rectify this by selecting a more retentive filter, such as a kitchen sieve with a tighter construction, capable of retaining smaller particles than the colander.

Although there are filters whose construction closely resembles that of a sieve, very few real-world filters are two-dimensional with a very narrow distribution or pore size.  Also, we should not ignore the fact that in most filtration processes the ‘pasta’ in our model would be seen as the ‘undesirable contaminant’, and the ‘cooking liquor’ as the valuable ‘filtrate’.  Accepting these observations, the inherent usefulness of our previous model now needs to be challenged; we need to add a degree of depth to our filtration model.  With this extra dimension, our separation no longer takes place on the surface of the filtration media and in a single plane but at some point in the journey of an individual particle through the thickness of the filtration media.

In our next blog we explore how inertial impaction works as a filtration mechanism.