Chasing coffee micro-fines (<10 µm) [Video & Pics & Updated] - Page 3

Grinders are one of the keys to exceptional espresso. Discuss them here.
samuellaw178
Team HB

Postby samuellaw178 » Mar 24, 2016, 7:07 pm

doublehelix wrote:In my hands adding filters sped up the extraction (espresso)--easy to imagine that adding additional filters would help trap more fines from winding up in your cup.


I've just tried this (with 4 Aeropress filters). As everyone has mentioned, adding filters speeds up the flow rate by a huge margin. Tasting the shot, the body isn't affected as much as I thought it would be (in agreement with Tony's experience, but keep in mind this is just one shot).

This could mean two things - either the aeropress pores size is too large to capture the micro-fines, or the micro-fines do not migrate into the cup(or it's not the significant mouthfeel component). I have an idea to test this though. Pull two shots (one with filters, one without). Then filter both via a 0.45 micron or even a 0.2micron filter. Compare if there is any significant difference in backflow resistance. From experience, filtering a normal shot would give a rather significant backflow resistance. If the filtered shots give little or no resistance, that'll mean the fines are removed but did not affect the body. If both gave the same resistance, then the aeropress filter is too large to filter the micro fines.

Also, the increased flow rate means that in the usual filter basket, the basket holes do experience clogging to a certain extent that artificially increases the back resistance.

User avatar
doublehelix

Postby doublehelix » replying to samuellaw178 » Mar 28, 2016, 11:46 pm

This is going to sound heretical. Socratic Coffe has an article in Standart magazine reporting on migration of finely ground coffee. Not much is noted in terms of fines migration.

One way to think about the aerpress filters is they act like a thin layer of packed coffee grounds, with the notable exception of presenting a fixed network geometry for water to pass through. The filter/ basket interface presents a type of stable plenum that attenuates the action of fines nearby basket holes from jamming water flow.

samuellaw178
Team HB

Postby samuellaw178 » Mar 29, 2016, 12:12 am

doublehelix wrote: This is going to sound heretical. Socratic Coffe has an article in Standart magazine reporting on migration of finely ground coffee. Not much is noted in terms of fines migration.


Is that article available online or shareable? Don't quite get what you mean by 'Not much is noted in terms of fines migration'. :oops: Do you mean they did not report much on the actual fines migration, or they did not see any fines migration during the brew process?

That's exactly how I imagine the filter to work as well. Paper filters are a essentially a network of cellulose fibres (a 3D filter if you will, compared to the single -plane filter holes on the basket).

User avatar
doublehelix

Postby doublehelix » replying to samuellaw178 » Mar 29, 2016, 12:54 pm

The holes in the metal basket are a bit different from paper. Paper presents a continuum of "holes" and obstacles, whereas the metal filter presents boundary conditions that truly impenetrable and well-defined. Water cannot pass through the spaces between the holes, whereas water molecules can permeate within and between cellulose fibers, albeit, in complex ways. There are filters that do look like a nano-version of a basket-- they used be called Nucleopore Membranes--https://en.wikipedia.org/wiki/Nucleopore_filter. These filters are plastic with well-characterized holes that look like they were made with a cookie-cutter.

Standart is just print..tests that SC did involved using extensive sieving to partition coffee particle sizes, then layering pucks with different combination of these size fraction. Frankly, one good way (need a real lab) to study fines migration is to label (fluorescent/radio-tagging sized particles), then run an extraction. Slice (section) the resulting puck like a salami and measured abundance of the label, will change upon migration, or remain static. Another way is to just measure the particle distribution, after extraction. Yet another way is to mix tiny, fluorescently labeled latex beads, of known size within the grounds before tamping and extraction, and then do the "salami" analysis. Of course one could also just measure what labels wind up in the cup.....etc....

My guess here is that folks involved in soil science know about really good assays and analyses that would reveal much about what is truly going on within a puck during extraction....all fun stuff to think about and try.

Patagent

Postby Patagent » Mar 29, 2016, 1:34 pm

doublehelix wrote:This is going to sound heretical. Socratic Coffe has an article in Standart magazine reporting on migration of finely ground coffee. Not much is noted in terms of fines migration.

One way to think about the aerpress filters is they act like a thin layer of packed coffee grounds, with the notable exception of presenting a fixed network geometry for water to pass through. The filter/ basket interface presents a type of stable plenum that attenuates the action of fines nearby basket holes from jamming water flow.


Just thinking out loud- is it the attenuation of "action of fines?" Sam mentioned earlier that the average pore size of a VST basket is 300 um. Based on the pictures, it look like fines are much much smaller than 300 um.

vit

Postby vit » Mar 29, 2016, 1:47 pm

Obviously holes in the basket are a kind of blocked by tamped coffee, so most fines are much smaller than diameter of the holes, limited by free space between the grounds / grounds and body of the basket. It also explains why the flow is much stronger if you put the paper between the basket and coffee

Patagent

Postby Patagent » replying to vit » Mar 29, 2016, 2:01 pm

I'm trying to understand how fines (at the interface of coffee and basket) affect flow rate. I guess what you're saying is that fines don't clog up the basket pore but rather the small tiny open spaces where the coffee grinds touch the basket? In sciencesque terms, the presence of the filter reduces time spent in the rate determining step.

samuellaw178
Team HB

Postby samuellaw178 » Mar 29, 2016, 11:29 pm

doublehelix wrote:Frankly, one good way (need a real lab) to study fines migration is to label (fluorescent/radio-tagging sized particles), then run an extraction. .


Thanks Dave!

I'm back with a bit more info, after a couple of hours spent. :? The good news is we probably don't need to go the fluorescence tagging route to confirm fines migration - unless it is to study the fines migration rate or to see it in real time.

So I pulled two espresso of 15g dose into 25g beverage. One of them is a normal espresso (pulled in 15g VST filter basket) ; the other is a 'filtered' espresso pulled with 2 layers of Aeropress filter in the VST filter basket. The filtered espresso has a slightly finer grind to prevent gusher. The samples were kept in a container for about 1 hour until they were viewed under the microscope. Multiple observations were done (shaked prior to sampling vs allowed to sit for 15-30 min and then sampled).


Here're some pictures to visualize what espresso fluid looks like at micron level (Yeah, these are the stuff that we stomach every day :? ):


Normal Espresso - unfiltered (without Aeropress paper filter):

100x magnification:

The brown grains are coffee grinds/solids
Image
Image

The reason for two magnifications - low magnification was to have a bigger view field (more representative of the overall sample), and high magnification was to look at the details (more details but less representative)


1000x magnification:
Image
Image
Image

Note: I had to vary the depth of focus in order to visualize either the emulsion droplets or the particles. In reality, they're probably suspended uniformly throughout the fluid.

Filtered Espresso - 2 layers of Aeropress paper filter

100x magnification
Image
Image

1000x magnification
Image
Image
Image
Image


From my limited observation above, fines migration definitely do occur (in both filtered and non-filtered espresso), and they are one of the major components in espresso. What was more surprising to me was how much oil droplets/emulsion are there in the cup.

Summary
The biggest differences between the filtered & non-filtered espresso are:
i) More emulsion droplets are observed in the filtered samples (I've got more pictures unposted but the trend seems true, at least for these samples) - not sure exactly why but we can always theorize.
ii) There are *slightly* less coffee grinds/solids in the filtered espresso, especially the larger grinds at >30um. Aeropress seems to be quite effective at filtering fines larger than 20-30 micron.

Other observations:
i) By taking the top part and bottom part of the unfiltered espresso, I can observe >30 micron grinds in the bottom part (due to sedimentation), but not in the top parts (suspended grinds are normally <20 micron).
ii) The edges of the coffee particles appear to be more rounded, compared to the initial pictures I've taken in the OP(which has sharp edges). This could be due to swelling in water (ethanol was used previously), or the pressure/flow in espresso has smoothen/eroded the edges of the coffee particles.

Because of the more rounded feature of the particles, I believe, but can't confirm, those micron-grains at 1000x magnification are indeed micro-fines . They could potentially be emulsion droplets (but unlikely).


Also, to confirm what I am seeing are actually real, I tried syringe filter with 0.22um(I believe that's what VST uses) and 0.45um membrane filter. I couldn't see anything under the microscope at all for both filters (except for some noises). No grains, no emulsion, nothing. The VST refractometer measures only the dissolved solids, but gives no information about the particulates/emulsion - nothing new but it's good to visualize the difference directly.

Syringe-filtered espresso - 0.22um filter

Image


I guess what would be real interesting to do next, is to visualize the shots from an EK43 grinder and a flat burrs grinder.

User avatar
doublehelix

Postby doublehelix » Mar 30, 2016, 11:37 am

samuellaw178 wrote:Thanks Dave!

I'm back with a bit more info, after a couple of hours spent. :? The good news is we probably don't need to go the fluorescence tagging route to confirm fines migration - unless it is to study the fines migration rate or to see it in real time.

So I pulled two espresso of 15g dose into 25g beverage. One of them is a normal espresso (pulled in 15g VST filter basket) ; the other is a 'filtered' espresso pulled with 2 layers of Aeropress filter in the VST filter basket. The filtered espresso has a slightly finer grind to prevent gusher. The samples were kept in a container for about 1 hour until they were viewed under the microscope. Multiple observations were done (shaked prior to sampling vs allowed to sit for 15-30 min and then sampled).





I guess what would be real interesting to do next, is to visualize the shots from an EK43 grinder and a flat burrs grinder.



Wow! Those are really neat pictures! In terms of fines migration--I think you've shown fines in the extraction (cup), but their origin is not clear. These could be fines that locally migrated from the first few millimeters above the basket/puck interface....?
Fascinating observations regrading coffee vesicles (emulsion).....in fact, extrusion through filters is a common method for preparing vesicles.

FWIW- here's a thread I started some months ago when I started exploring the filter/basket interface: AeroPress meets espresso machine