Bimodal vs. unimodal particle distribution: what really lies behind syrupy mouthfeel
Bimodal vs. unimodal particle distribution: what really lies behind syrupy mouthfeel
"Unimodal" is one of the best-selling words in grinder marketing. We went and looked at what the measurements actually show – and learned along the way that the thing itself doesn't exist. What does exist is a clean physical explanation for what we experience as body.
Every espresso grind is bimodal, often even trimodal. A unimodal espresso distribution has never been measured in the scientific literature – the correct term would be "fines-reduced", not "unimodal".
Fines fill the gaps between the larger particles, lower permeability and lengthen extraction. That is well documented. That they create body is not.
Syrupy mouthfeel, as far as we know today, comes from several phases at once: dissolved solids, emulsified oil droplets, suspended cell wall fragments and the foam. It is not a pure viscosity question.
What bimodal actually means
Bimodal means: the particle size distribution has two maxima. One peak at the finest particles, the so-called fines, and one at the coarse fraction. Conventionally, particles below 100 micrometres count as fines – that is an agreement, not a constant of nature.
A asks: how much coffee mass is in this size? Two mountains — fines and coarse fraction. B asks: how many grains have this size? Only one mountain left, the coarse fraction almost disappears.
Curve shape schematic, positions from measurements: fines peak 40–50 µm (Corrochano et al. 2015), peaks 13 µm and 257 µm (Uman et al. 2016), 99 % of particles below 100 µm number-weighted (Cameron et al. 2020).
Both pictures show exactly the same grind. Above, mass counts; below, the number of grains – and two mountains have become one. This is exactly where most forum misunderstandings start: two people talk about the same grinder and mean two different charts. The orange curve above is what fines-reduced burr geometries achieve: a smaller fines peak. It doesn't disappear. Analysing 300 distributions, Gagné found that the position of the fines peak varies between different grinders by up to 8 micrometres – real, but small.
Peer-ReviewA warning about reading the numbers, one that almost led us astray ourselves: the widely quoted line "99 percent of all particles are smaller than 100 micrometres" from Cameron et al. (2020) is number-weighted, not mass-weighted. By mass, fines are a small minority. They do, however, account for around 80 percent of the accessible surface area. And: different measurement methods – laser diffraction, dynamic image analysis, sieving – return different numbers. Values from two studies cannot be compared directly.
Why unimodal espresso doesn't exist
This is the point where we had to rethink the most while researching. In all the peer-reviewed literature we reviewed there is not a single measured unimodal particle distribution at espresso grind size. All of them are bimodal, and Gagné's analysis of 300 distributions shows that the data are described even better by three overlapping log-normal distributions than by two.
The mechanism behind it is in Cameron et al. (2020): when the grind is set finer, the proportion of fines rises – their size stays constant. Particles keep breaking until they are small enough to pass the burr gap. The fines peak is a product of fracture mechanics, not of gap width. Grinding finer creates more fines, not fewer.
The only experimentally documented narrowing of the distribution comes from a different direction: Uman et al. ground beans at minus 196 degrees in liquid nitrogen. The modal value dropped by 31 percent, the fines cutoff from 70 to 61 micrometres. Narrower yes – unimodal no.
If someone sells you a unimodal grinder, what they factually mean is: the fines peak is smaller. That is a real, measurable property and well worth something. It just isn't what the word claims. That is why we prefer "fines-reduced" – it describes what actually happens. Others in the industry take it more lightly, and that is fine too.
Fines, permeability and the extraction curve
Here the evidence gets really good. Corrochano et al. (2015) measured the permeability of coffee beds systematically: it ranges from 2.59 × 10−14 to 3.36 × 10−13 square metres – a factor of around 13 between the extremes, from particle distribution and bed density alone. At the same bulk density, a coarsely ground bed was 4.4 times more permeable than a fine one.
The most revealing value: the measured bed porosity lay between 0.12 and 0.33 – clearly below 0.36, the value for a random packing of equally sized spheres. That is the direct physical evidence that fines fill the gaps between the coarse particles.
On the left, water finds continuous channels. On the right, the fines settle into the gaps and narrow every path.
The bars show the extremes of the measured permeability: 3.36 × 10⁻¹³ against 2.59 × 10⁻¹⁴ m², a factor of around 13 (Corrochano et al. 2015). Particle sizes greatly exaggerated in the image — real fines are about ten times smaller than drawn here.
Smrke, Eiermann and Yeretzian (ZHAW Wädenswil) put the same finding plainly in 2024: a higher fines content lowers permeability, slows the flow and lengthens extraction time.
What does not follow from this is "more fines equals more extraction". Cameron et al. (2020) found a non-monotonic curve: extraction yield rises with a finer grind only up to a point and then falls again. The reason is agglomeration and uneven bed density, which partly block the flow. The gap between calculated and actually measured yield was 13.1 percent at the finest setting – the share of the bed that water no longer reaches by calculation. At slightly coarser settings this gap shrinks to 6.1 and 2.6 percent. So the effect is not a law of nature but a consequence of grinding too fine.
The authors' conclusion is particularly instructive: the same yield of 22 percent can be reached with two completely different grind settings – but the chemical composition of the two shots is not the same. Two espressos with identical refractometer readings can taste entirely different. In the grinder simulator you can play this behaviour through.
What syrupy body physically consists of
Now to the actual question. In 2011, Illy and Navarini described in Food Biophysics what an espresso consists of as a physical system – and the numbers are remarkably concrete:
| Phase | Value |
|---|---|
| Dissolved solids | around 52.5 g/l |
| Emulsified oil droplets | 90 % smaller than 10 µm, volume fraction 0.2–0.3 % |
| Suspended cell wall fragments | 2–5 µm, around 150 mg per cup, which is about 5 g/l in a 30 ml cup |
| Total lipids | Arabica 1.80 mg/ml, Robusta 0.55 mg/ml |
| Crema | foam density 0.40–0.60 g/ml, bubbles 10–150 µm |
What surprised us during the research: we found no published measurement of the shear viscosity of espresso. Illy and Navarini, like Ferrari et al., deal with interfacial properties, not with the viscosity of the drink. Every concrete millipascal-second figure you read about espresso is, to the best of our knowledge, unsupported.
That argues for a particular reading: in terms of concentration, the dissolved solids alone sit below the threshold at which people can reliably detect differences in viscosity at all. What carries the syrupy impression is probably the dispersed phases – oil droplets, microparticles, foam – plus the cross-modal coupling to sweetness and aroma that we described on the sensory page. That is an argument, not a proof, and we are deliberately labelling it as such.
What this means for choosing a grinder
The common dichotomy – conical equals bimodal equals full-bodied, flat equals unimodal equals clean – does not hold up to measurement. Gagné did find that conical burrs are on average less unimodal and less even, but with a lot of overlap: you can easily find a flat burr set that is less unimodal than a conical one. And one finding contradicts the usual narrative directly: unimodality and evenness correlate positively. You are not trading one for the other.
Another finding that takes the wind out of marketing: Smrke et al. (2024) varied the fines content systematically and had the results assessed sensorially. Result: no sensory disadvantage from a higher fines content. That puts the whole "fines are bad" narrative on thin ice – and equally the counterclaim that fines guarantee body. We simply don't know precisely yet.
In practice, the way we see it, that means: choose the grinder by what is measurable and noticeable – evenness, retention, handling, grind adjustment resolution, burr size. How burr size plays into it we described under grinder burr sizes explained, and the comparison of the top models under Weber EG-1 vs. Lagom P80 vs. Lagom 01. You will find the whole range at coffee grinders.
Frequently asked questions
What is the difference between bimodal and unimodal particle distribution?
Bimodal means two maxima: a fines peak and a coarse fraction peak. Unimodal would be only one maximum. In the peer-reviewed literature a unimodal distribution has never been measured at espresso grind size – all of them are bimodal or even trimodal. The correct term for grinders with a small fines peak is "fines-reduced".
Do fines create more body in espresso?
What is documented is that fines lower permeability, slow the flow and lengthen extraction time. There is no causal proof that they create more body. Smrke et al. (2024) found neither a sensory disadvantage nor a demonstrated body advantage at a higher fines content.
Where does syrupy mouthfeel in espresso come from?
As far as we know today, from several phases at once: around 52.5 g/l dissolved solids, emulsified oil droplets below 10 µm, around 150 mg of suspended cell wall fragments per cup and the crema foam with a density of 0.40 to 0.60 g/ml. We could not find a published viscosity measurement of espresso.
Are conical burrs more full-bodied than flat ones?
As a general rule that does not hold up to measurement. In an analysis of 300 particle distributions across 24 espresso grinders, conical burrs were on average less unimodal, but the overlap between the designs was large. The design alone says little about body.
Does bean origin influence particle distribution?
No. Uman et al. (2016) tested coffees from Guatemala, El Salvador, Tanzania and Ethiopia by analysis of variance: the differences were not significant. What clearly changes the distribution is the bean temperature during grinding – beans ground cold give a narrower distribution.
Roastery tip
If you want to taste the difference between fines-rich and fines-reduced, take a coffee with plenty of structure. Cozy Chocolate shows the dense, round side, Wild Peach the clear, fruity one. The same coffee on two different grinders opens more eyes than any curve – and that is the only measurement that counts in the end. Our recommendation, one of many.
