The taste of fines (coffee grind)

The smallest particles in coffee grind are called “fines”. The role of fines for the taste is disputed – are they god or bad for the taste ?

Note: This post looks mostly on drip brewing but the findings can be used for espresso and immersion too.

How large are fines ?
There is no common size definition.
Randy Pope from Bunn talkes about below 600 μm
Matt Perger talkes about a 100 μm limit, but started off using a 250 um limit.
Note: 100 μm (microns or micrometer) equals 0,1 mm.

 

If you brew a very fine grind it will taste bitter and have other over-extracted tastes. So you should think that removing fines in a grind would result in a better taste.

But its not that simple. Read here how Matt Perger first sifted his coffee for a world competition to get rid of the fines. But then later found that great-tasting-grind has a lot of fines: https://baristahustle.com/blogs/barista-hustle/a-wide-thought-distribution-about-grinding

cup-only-22majSo how come fines can both appear to give bitterness and not to ?

It has long puzzled me how small particles in the coffee grind could work so different for taste. I hold a Master degree in Food Science and wonder about things like this. I had a suspicion that the reason had to do with how the fines were “situated” in the grind.

The coffee particles are not all free floating among each other (like often portrayed in graphic illustrations of a grind). The fines are sticky. That’s due to the physical forces called adhesion. Like when flour stick on your clothes.

The fines can either stick to larger particles – or to other fines, see photos below. My conclusion is that each result in a different taste:

When fines are attached to bigger particles they act as part of the large particle in the extraction and do not over-extract as when they are separate where fines only stick to other fines.

Here come some experiments that supports this theory. But first some microscope pictures where you can see how the fines stick.

Photo 1: normal grind, untouched
Medium roasted Kenya coffee
Grind untouched

Here clearly the small particles are attached to the larger particles (could be called agglomeration).

The particles stick together but are not strongly bonded. On the other hand you can’t just blow them off. They more stick like flour on your clothes. You can rub them off -> see the sifted grind below.

Also note in the lower right corner there is small particle – looks like several fines that stick together. That’s what I here call “separate fines”.

Photo 2: sifted grind
Grind sifted 400 to 1100 um

The same grind sifted with a simple Kruve sifter. This is the fraction between 400 and 1100 μm. Most fines on the surface are gone. But still a few stick.

When you shake with a larger sifter (like 1100 μm) the grind quickly reaches a steady-state = the amount staying back does not change (within 30 seconds). Where as with the smaller sifters (like 400 μm) you can go on for a several minutes and keep getting more small particles through. That is because of the adhesion forces: it takes some work to loosen the fines from the bigger particles.

.
Photo 3: the boulders fraction, above 1100 μm
Grind over 1100 um chaff

The largest particles of the grind, called boulders. With a “golden” chaff as well.

 

Photo 4: the fines fraction, below 400 μm

Grind under 400 um

The fines stick together even though the grind has gone through two sieves; the 1100 μm and the 400 μm. They were probably separated during the shaking, but hooked up again.

Small particles just have an urge to stick. Maybe it due to surface tension like water. When it rains on a window and a drop runs down the vertical surface – a long trace of water stays back. But within seconds it gathers in drops instead of being spread out. Thats because the surface tensions is lower in a round drop.

 

Before I get to how these four grinds taste (see Experiment 2) then first these questions:

Won’t the fines just get flushed off the larger particles when water is added in the brewing process – so it won’t matter how they were situated in the dry grind ? Well, several taste experiments suggestes that is not the case.  See the taste experiments below.

 And; do fines ever appear separate in daily life like on photo 4 ?

This investigation started off by a discovery by both Andreas and me. Independently we found that grind from a bulk grinder (big burrs and high speed grams pr minute) tasted better after reducing visible fines. Less bitterness. This was for filter brewing (Kalita and V60).

How we reduced visible fines:
1) At the end of grinding don’t shake/knock the grinder to loosen the fines from the outlet to get all of it for your brew. Instead remove the container with the grind and then cleanse the grinder by shake/knock/brushing so it’s clean for the next batch.

2) In the container you grind into – remove the fines that stick to the edges (either with a finger or a spoon). The light parts are chaff but dark fines are in there too.

Thereses EK ned i pitcher

This is Taste experiment 1 that shows the effect of the separate fines on the taste. When you remove them the brew taste better.

Later, after all these investigations I have found this procedure to visible fines: Stir the dry grind with a spoon before placing it in the filter – with the intention to get the separated fines to blend in and stick to the larger particles.

Remember; the fines are not strongly bonded to other particles. They are easily loosend but also quickly stick to something else again.

Here is a grind from a fast bulk grinder. Look at the grind pile; at the top you see fine particles. At the bottom you see coarser particles. Here I would remove the fines on the edge and the stirre the grind pile it self.

Fil 06-08-2017 21.27.53

Taste experiment 2

I have made a taste comparison of the four grinds from the microscope photo above – made as cupping. Tasted when cooled to around 55°C.

Cup 1, untouched grind as on photo 1:
The best tasting of the four cups. Good body, pleasant acidity not too much, and the aroma characteristic of the coffee.

Cup 2, sifted grind, the middle fraction between 400 and 1100 μm, as on photo 2:
Clear, thin, acidic, dry and some bad aftertaste.

Cup 3, sifted grind, “boulders”, fraction larger than 1100 μm, as on photo 3:
Thin, clear, acidic – but with when cooled further the acidity disappeared.

Cup 4, sifted grind, “fines”, fraction smaller than 400 μm, as on photo 4:
Bitter and other over-extracted tastes (not pleasant).

I tried to mix the 3 fractions: one spoonful of the fines fraction, four spoonful of the middle fraction and one of the boulders fraction. It didn’t taste like the normal/untouched grind. It had a sharp/tart taste. And still not the pleasant aroma found from the untouched grind

Conclusion
Fines don’t act on their own in the brewing process. It does make a difference whether they are attached to the larger particles – or are separately only attached to other fines.

Experiment 3 – drip/filter brew with separated fines

Comparing normal grind with fines attached to larger particles versus the fines separated out by sifting.

First brew: untouched grind at a grind size that gives a good extraction on drip/filter, like V60 or Kalita

Second brew: a grind at same grind size, just sifted first to separate the fines. Both fractions, the fines and the rest of the grind is placed in the filter. Not stired because you want to keep the fines separate. Brewed the same way as first brew.

Photo 5 and 6:

Fil 07-08-2017 18.11.30Fil 07-08-2017 18.11.46

Here I sifted 25 grams with the standard-method I use here: 400 and 1100 um size sifters. (Not that the >1100 um fraction is needed here. It was just to repeat the way the middle-fraction 400-1100 um is made.) I placed the fines fraction first in the filter so they wouldn’t mix up too much with the larger particles and stick back on.

Fil 07-08-2017 18.12.00

Taste
The brew with separate fines tasted sharp and had bitterness. Compared with the “untouched” grind it had more acidity, less body and less aroma.

Mikkel did the exact same experiment and got the same impact on the taste.

Conclusion
Since the ratio of fines is the same in the two brews – this can only be explained by a difference in how the fines are placed; only attached to other fines – or attached to bigger particles.

Microscopy of wet grind after extraction

I also did microscopy on grind after brewing to compare sifted grind i.e. particles without fines attached – versus normal grind with fines attached.

But it wasn’t as clear as with the dry grind (not this far anyway, maybe it can be done better). Because with the water the particles swell up and kind of integrate with the fines.

Here microscopy of a normal grind after a V60 extraction. You can kind of make out how small particles are attached to bigger one. But they don’t stand out as clear as in the dry grind.

Photo 7-9: “Untouched” grind

Vario untouched grind 1 from V60Vario untouched grind 2 from V60Ek grind untouched2 V60

Photo 10: sifted grind, the 400 – 1100 μm fraction

Vario sifted 400 to 1100 um V60The particles look more naked. But it’s not that clear.

 

What I found the most interesting of the wet grinds so far: is to see what the separated fines looks like. They look like totally dissolved into a gel, no longer a dark cellulose core:

Photo 11: fines fraction – from a thin “mud layer” on top of the grindVario grind top mud layer V60

Photo 12
Cupping grind u400um 1

This is the fraction from the Kruve sifter with particles below 400 um.

I made it by places the grind in a cup and then just enough hot water to mix it up (like 1-2 cm). And then let it sit for a while. Afterwards I wondered if cool so quickly that particles didn’t swell up as much as in the brew process.

Anyway, for what it is: here is a spectrum of the smaller particles. Clearly a difference between those around 300-400 um and the smallest, probably below 100 um. The smallest looks a somewhat disolved.

Reflections from this

Reflection 1
Maybe this also illustrate the point of bloom in the beginning of the brew – where just a little water is added to the grind and then let it swell up for 30 seconds. Because here the grind can form this gel-body where fines and larger particles integrate (maybe into a gellyfish-type body ?)

Reflection 2
In the last part of coffee brewing, like when the last water draws down in the filter on V60 or kalita – if it is shaken it makes the coffee over-extract. That is; after done pouring the water and the grind just slowly falls to the bottom.

Seen in this light with the importance fines covering the particles … when pouring the water into the sludge these lumps float freely around in the turmoil … But when you stop pouring water the turbulence of the sludge slows down … and the grind settles on the bottom … Maybe when the particles settle on top of each other and then get shaken (like by the Rao spin) … then they will rub on each other and scrub off the loosely attached fines .. resulting in over-extraction.

Reflection 4
This makes me speculate: could it be the “dark cellulose part” that holds back the unwanted over-extracted taste ? So it’s only when we got these fines-to-fines alone that we get the bad over-extracted taste ? Just speculation.

Reflection 3
If the fines do form a jello/gel this could explain why an espresso stall when the grind is too fine. Stall = the water can’t go through/the puck is clogged = nothings comes in the cup.

Could be interesting to see …
It would really be interesting to see what’s going on during the brewing process. It’s just difficult to get under the microscope where you have the best view when the grind is spread out on a white background. The temperature would not be as high as when brewing (it quickly drops when spread out). And I have the feeling that the closeness of the grind is important to how the fines connect with the larger particles.

What to do ?

If your grind has visible fines you can:

1) try to remove visible fines aggregation

2) stir the dry grind before starting the brew process – in order to get the fines attached on the larger particles.

 

Grind with visible fines:

Fil 06-08-2017 21.27.53