Chasing the Peak: A Dimensional Matter

Experience shows that espresso coffee really sings when you can extract high concentration of solubles over a relatively short time. Achieving this is all about increasing the contact surface between the coffee and water. This helps transfer soluble and emulsifiable materials into the brew. Chemically, the finer the particles, the quicker the extraction. However, there is also a physical/mechanical need for the water to flow through the ground coffee bed, which becomes difficult with an overly fine grind. Therefore, for optimal extraction, both fine and coarse particles are needed, at least when there is no material other than coffee present.*

This is where particle size distribution comes into play. The particle size of coffee grinds influences many properties of the resulting beverage. The size and shape of powders affect how the ground coffee compacts in the basket upon tamping, how they extract, dissolve, shape the viscosity of the liquid coffee and ultimately, its taste.

Most particle size analysis techniques assume that all particles are spheres. Of course this is simplistic, as we know that coffee particles come in all shapes and sizes. But, this assumption is generally not too far off for coffee—and it makes calculations and data interpretation much easier.

Particle size data is generally represented in three main ways. These describe how particles of a certain diameter contribute to the overall:

  • Number of Particles

  • Surface Area

  • Volume

These three representations can result in graphs that look strikingly different, while they represent the exact same data.

Particle number contribution is easiest to visualise. For example, we look at coffee grinds through a microscope and simply measure the dimension of each particle—counting up the number of small ones versus the big ones. As this only takes into account the diameter (d) of a certain particle, it’s a one-dimensional (1D) measurement. Number distribution is often the kind of data spat out by particle sizers using the laser diffraction method. 

 When we look at the contribution a certain particle makes to the overall surface area of the whole powder, we move from 1D to 2D. The surface area contribution is proportional to the diameter of the particle squared.

Similarly, when we want to know how a particle contributes to the overall powder volume, we need to look at the particles in 3D. The volume of the spherical particle is proportional to its diameter cubed. Since volume and mass are directly proportional—this volume measurement is actually a proxy for weight. Volume/weight distribution is the kind of data you get from a sieving measurement.

Imagine a coffee grind as a sphere. Now we can calculate the surface and volume of each particle.

Coffee is a hugely varied powder mixture in terms of size, containing very fine particles from a couple of microns (a millionth of a meter) up to 1-2 millimeters. That means a 100-1000 times difference in scale between the smallest and largest particles. Imagine that squared or cubed…! For example, a particle 1 mm in diameter has a surface area ten thousand times larger than one measuring 10 micrometers across. Its volume would be a staggering one million times that of the smaller particle! Here’s a zoomed out electron microscope image once again to show this great spread of particle size:

From tiny fine particles to big bits: SEM image of coffee grind at 200x magnification.

No surprise this will mean a massive difference when it comes to how we read grind particle distribution data. Let’s look at three different representations of particle size distribution data gained from the same (imaginary) powder below.

From the first, numerical representation in the top panel, we can gather that there are many fine particles, a significant amount of medium particles, and practically no large particles in the mixture.

Moving from 1D to 3D: Number, surface area and volume contributions to the grind particle size distribution of the same sample powder

The second panel shows how the same particles contribute to the overall surface of the particle mixture. It’s obvious that, even though there is a large number of fines, they contribute very little to the overall surface area. The medium particles now appear to be dominant, representing ~75% of the overall surface area. Large particles suddenly appear on the surface area graph, showing that even just a couple of large particles contribute nearly 25% of the overall surface.

The difference is even more striking in the bottom panel showing volume contributions of the same particles. The fines, despite their large number, represent a very small, practically negligible volume compared to the medium and large particles. The large particles, even though there are very few of them, contribute about two thirds of the overall volume. Which also means two thirds of the overall weight is made up of these large particles—how could you have known that just looking at the 1D graph?

This is such an important point, I’m going to say it again. These three graphs all show the exact same data, plotted in three different ways.

This means that we really have to keep in mind which representation we look at when interpreting grind distribution data! If you look at a grind particle distribution graph next time, make sure to check out what’s exactly shown on the y axis. As we have shown, the exact same sample can look strikingly different depending on the way the data is represented, which can be quite misleading. (Don't worry, it only takes a little mathematical manipulation to convert between these three).

As I mentioned at the top of this article, coffee extraction is strongly related to the contact area between the water and the coffee (further theoretical calculations showing this can be supplied on request). This means the surface area distribution graph is the most meaningful when considering how particles of different diameters contribute to extraction.

Bottom line: make sure you pay attention to the surface area graph!

One last point: don’t confuse this with an increase in the overall surface area upon grinding. When a big particle is broken into smaller pieces, the overall surface area increases, which is why we grind coffee in the first place. All of the above relates to distribution within a finished grind with a given overall surface area.

Next time we will look at other common pitfalls of interpreting grind particle size distribution graphs and give some more tips on how to make sure you get the best out of your grinder! Stay tuned!

*(or you could mix it up and try this)

SUBSCRIBE to Coffee Science Lab to get the latest coffee science updates, or follow us on Instagram!

Tags: #coffee #science #coffeesciencelab #surfacearea #coffeegrinds

150 views0 comments

Recent Posts

See All