In the previous blog post of this series I looked at the chemical compounds present in a cup of coffee and how the concentrations span 9 orders of magnitude. Each of the compounds present in coffee will have a specific solubility and extraction rate. This sounds bad, but is actually good, because roasted coffee is not necessarily “perfect”, even when roasted to perfection. If all compounds were extracted at the same rate there wouldn’t be any room for creativity and variation when brewing coffee. Luckily this is not the case. If the goal with a coffee extraction is to most accurately reflect the soluble part of a coffee bean you should grind as fine as possible and steep for a long time to ensure that also the slower extracting compounds are extracted. To speed up matters you could also boil the coffee during extraction. The result would be similar to a Turkish coffee brewed without addition of sugar or spices. The extraction yield would approach 30%. The result could be strong and tasty, but it’s not necessarily the best coffee you can brew. To achieve this you will most likely want to do a selective extraction. I sometimes see the term “even extraction” used to describe a good coffee, but from a chemical point of view what you really are aiming for is an uneven extraction where compounds have not been extracted to the same degree. The best way to achieve this is with the pour over or percolation technique (in the following I will stick with the term pour over). Percolation in this context BTW should not be confused with a coffee percolator where convection causes the heated coffee to be lifted in a central tube so that it can trickle back down through the coffee bed. The first models were invented more than 200 years ago, and updated versions became very popular in the 70’s.
Pour over brewing may sound deceptively simple: You pour hot water over ground coffee and let the brew drip into your cup by gravity filtration through a filter. Once you think about it though there are surprisingly many parameters that you can adjust. In part 1 I discussed how brew ratio, grind size/particle size distribution, water composition and temperature, filter material and pore size all affect the end result. But there is more – much more! The following also affect the end result:
- the geometry and size of the filter holder
- how fast and at what intervals the water is added (this includes whether a bloom phase where coffee grounds are allowed to swell a little and where trapped CO2 may be released)
- the degree of agitation (mechanical with a spoon, a gentle swirl, or by pouring water sufficiently fast to create turbulence)
- brew outflow (controlled by valve?)
- to which degree the water is forced through the entire coffee bed or is allowed to bypass the coffee grounds, either via channeling or simply by exiting through the filter paper as soon as possible
An additional complexity arises because many of the parameters are correlated. In the table below I’ve tried to indicate the correlations. I have categorized the parameters you can influence directly as independent. The rest are categorized as dependent.
As an example we can look at grind size and particle size distribution (PSD). The most important effect is that the total surface area available affects the extraction. Finer particles will slow down flow and thereby increase contact time (and hence improve extraction). There is also a risk of filter clogging which can further slow down the outflow. It is not unreasonable to think that grind size and PSD can influence channeling, but exactly how is not obvious to me. With a fine grind size and slow flow any water standing above the coffee bed may just as well take the shortcut through the filter and bypass the entire filter bed, because water is lazy. No matter how much you would like for the water to follow a certain path in order to dissolve and extract all the good stuff from the coffee beans, it will always take the path of least resistance. The amount of bypass will increase with a finer grind and a filter holder with a pattern that reduces filter contact, thereby increasing the flow rate of water through the sides of the filter.
We could go on like this for all the correlations shown in the table above, but I’ll spare you for that. I hope you agree with me that pour over brewing in fact is an immensely complex system to accurately describe and control. And I think the countless YouTube videos on pour over brewing techniques testify to this. Nevertheless, the pour over method is popular because it gives professional and home baristas alike a lot of control over the final outcome.
The downside of this complexity is that it introduces unintended and random variations which in turn may lead to coffee preparation myths. A further effect is that two persons following the same “recipe” will not necessarily obtain similar coffee brews, simply because most coffee recipes fail to describe all the parameters mentioned above with sufficient detail.
Am I overthinking it? Maybe, maybe not. As I began to read up on drip coffee I was amazed by the amount of thinking and innovation within the field of drip coffee makers. And it turns out people have been thinking about coffee makers for very many years! Just take a look at the picture below from the book “All about coffee” published in 1922.
Out of curiosity I then began to collect and systemize some of the coffee makers available today for pour over brewing. The endless list (see table below) may almost seem silly at first glance, but many of the small differences are actually there for a purpose. The inventors behind many of these devices are really trying to untie the coffee knot. I’ve recorded the following data about the coffee makers:
Material and insulation: When brewing coffee you want to maintain a relatively high temperature. Metal is bad since it is a good heat conductor, unless you have a double walled filter holder with vacuum such as the Stagg X/XF or there is so much metal that the thermal mass is significant compared to the volume of water used for brewing (the only example I can really think of in this category is an all metal espresso portafilter). Vacuum insulated filter holders are also available in glass, but in my opinion a filter holder is handled so much that glass seems too fragile – sooner or later it will break. Personally I would prefer plastic (such as the red Hario V60 in the top picture) or ceramic based on cost, limited heat loss and sturdiness.
Shape/geometry: Many will be familiar with the Melitta type trapezoid shaped filter holder. In recent years cone shaped filters such as the Hario V60 have become popular, and then there are the flat bottom filter holders such as Kalita Wave that use folded/W shaped filters. The shape and size of the coffee bed differs between these designs, and it will influence how water flows through the coffee. A high/narrow cylindrical coffee bed will ensure good contact between coffee grounds and water, but the flow will be slow. A flat coffee bed on the other hand ensures a faster flow, but the water may quickly find an easy way out (channeling).
Inside wall surface: If a plain/smooth filter is used in a tapered or cone shaped filter holder there is a high risk that the wet paper will stick to the surface if the surface is flat, thereby effectively blocking the flow. A prime example of this is the Chemex. This may sound like a disadvantage at first, but the benefit is that there is essentially no bypass since the flow through the paper on the sides has been blocked. Most other setups either use a fluted/folded filter paper or introduce ridges or a pattern to allow a flow of coffee through the filter, even when it clings to the surface.
Filter type: Some brewers come equipped with metal filters out of the box, but the most common filter material by far is paper. Cylindrical brewers such as the Aeropress will normally just have a plain filter paper disc. Cone and trapezoid shaped filter holders will normally have a filter paper that is tailored to fit the shape of the filter holder. These filters are normally plain/unfolded as well. A challenge with filter paper is that coffee fines can clog them, resulting in a complete halt of the filtration. One way around this is to increase surface area, and this can be done by using a fluted or W-folded filter. The Kalita Wave filter is perhaps the best known example using a W-folded filter paper. In chemical laboratories fluted filter paper is used all the time for rapid gravity filtrations in glass shaped conical funnels. The Karlsbader coffee brewer is an exception to all of the above, because it only features a coarse ceramic grating as filter.
Inflow control: In the simplest designs you simply pour water on top of the coffee bed. The drawback is that this disturbs the coffee layer and redistributes the grounds which may create paths of less resistance (channeling). This can be prevented by controlling the inflow. One way is to reduce the pouring height and the diameter of the spigot. The characteristic goose neck kettles have been specifically designed for this purpose. A more fool proof solution might be to use a special water distributor on top of the filter holder such as the Gabi Dripmaster B or the Melodrip. It features a number of small holes which ensure a slow even flow of water onto the coffee bed. These will ensure a perfectly flat coffee bed. A drawback however of many water distributors (especially the uncovered ones) is the temperature loss.
Exit holes and outflow control: Besides grind size, geometry/surface and filter paper, the exit holes also influence the overall flow rate. Filter holders designed to maximize flow feature a single large open hole in combination with marked ridges or a patterned surface. Smaller/fewer holes and a plain surface which allows the filter paper to cling to the surface can both contribute to a reduced flow. Some of the filter holders feature an additional valve which can be used to restrict or even completely shut off the flow. This is a good idea as it allows for better control of the extraction time. Starting an extraction with the valve closed allows for an initial immersion or steeping phase, followed by a final draw down or percolation phase. The best known example here is probably the Clever brewer. A shut off valve also makes it easier to scale down a recipe, because less coffee will inevitably result in a faster flow. You could of course grind finer, but then you may also have to adjust the brew ratio, which again influences flow. See why pour over brewing is complicated? A valve solves this by allowing you to get the desired contact time.
Bypass: Water that bypasses the coffee bed does not take part in the extraction and only serves to dilute your coffee. This is OK if the coffee would otherwise be too strong, but generally you would want to avoid or reduce bypass as much as possible. This is also what Jonathan Gagné has arrived at in his recent blog post on optimal percolation brewing. A plain filter paper clinging to a smooth surface in a filter holder is a good way to reduce or even completely shut off bypass. How fast you add water will also influence bypass. You can minimize bypass by adding the water so slowly that it doesn’t stand above the coffee bed. If water stands high above the coffee bed the easiest way out is through the sides, rather than through the coffee bed. Ridges or a patterned surface on the filter holder as well as a fluted/folded paper will further contribute to this bypass. I’m tempted to say that many coffee makers are designed to allow for bypass, as this is a simple way of ensuring a certain flow and avoiding clogging. It renders the coffee maker a little forgiving if there are too many fines. The drawback is that the bypass is uncontrolled. It happens, but it’s not easy to observe to which degree. The best way to completely avoid bypass is to use a filter holder where the entire filter is located below the coffee. From an extraction viewpoint I would argue that as far as possible, only the coffee should restrict the flow of water – not the equipment (the exception being drippers that feature an on/off valve). What puzzles me is that there are not more cylindrical filter holders on the market with a flat filter situated entirely below the coffee bed. The best known example is still the Aeropress but there are other noteworthy examples as well: Delter, Tricolate, Büchner funnel, or the Proper and the Vietnamese Phin – to mention a few). In the table below I’ve done a quick assessment of the different filter holders and whether they allow any significant bypass or not.
Most of the equipment listed in the table below is for pour over coffee, but for comparison and benchmarking I’ve also included some other common coffee makers. Click the headers to sort the table on different columns. On mobile devices it may be easier to view the table as a static image.
* = accessory to filter holder
|Aeropress + Gabi dripmaster B||P||M||S||N||P||F97||N||Y||P|
|April Pour Over Brewer||C||M||T||R||F||FRO||Y||N||N|
|Blue Bottle Dripper||C||M||T||R||F||FR1||Y||N||N|
|Bonavita immersion dripper||C||M||T||R||P||TR1||Y||N||S|
|Brewista flat bottom||G||M||T||N||F||FV3||Y||N||N|
|Brewista flat bottom steeping||P||M||T||R||F||FR1||Y||N||S|
|Brewista full cone||G||M||T||R||P||CO||Y||N||N|
|Brewista tornado duo||G||V||T||R||P||CO||Y||N||N|
|Delter Coffee Press||P||M||S||N||P||F54||N||Y||P|
|Electric dripper (programmable)||P||M||T||R||P||TR1||N||Y||Y|
|Electric dripper (simple)||P||M||T||R||P||TR1||Y||N||N|
|Fellow Prismo for Aeropress||P||M||–||–||M||FP||N||N||Y|
|French Press (generic)||G||M||S||N||M||FP||N||N||P|
|Gabi drip master A||P||M||T||N||F||FR1||Y||Y||N|
|Gabi dripmaster B *||P||M||–||–||–||F33||–||Y||–|
|Karlsbad porcelain (generic)||C||M||S||N||–||FP||N||Y||N|
|Melitta filter holder||P||M||T||R||P||TR1||Y||N||N|
|Neapolitan flip pot||M||–||S||N||M||FP||N||N||N|
|OXO Pour-Over Coffee Maker w/water tank||P||M||T||R||P||TR1||Y||Y||N|
|Stanley pour over||M||–||S||N||M||CVP||Y||N||N|
|Swiss gold KF250||P||M||S||N||M||F||N||Y||N|
|Technivorm brew basket||P||M||T||R||P||TR1||Y||N||S|
|Wire filter holder||M||–||T||–||P||CO||Y||N||N|
If we leave the “ideal” extraction and the challenges of pour over coffee aside for a moment: Is there anything else we would want from a coffee maker and coffee preparation method? Yes indeed! I would put repeatability on top of my list. A foolproof method that doesn’t require too much attention. This is where immersion brewing enters the scene. What I like as a chemist with immersion brewing is that it’s a robust method. Share the recipe with two people and they will actually brew very similar coffee. It is also a method that doesn’t require my full attention all the time.
In immersion brewing the coffee grounds are steeped in water for a given amount of time. All the water is normally added at once. The slurry is stirred initially to help get rid of gas bubbles. After a given amount of time the resulting coffee brew is separated from the grounds by filtration. Pressure filtrations are the most common and include examples such as French press (and Aeropress if using the inverted method). A benefit of pressure filtration is that brewing time can be controlled more or less independently of the grind size. How long you leave the grounds to be extracted by the water translates into a given amount of total dissolved solids (TDS) and extraction yield (EY). Both TDS and EY will increase with time and eventually level out. At some point one will approach an equilibrium between the coffee brew and the coffee grounds. This is when the extracted coffee begins to resembles the composition of the soluble part of the coffee bean. The grind size however is still much coarser than what you would use for the “extract all” technique I mentioned in the beginning, and the temperature in immersion brewing quickly goes below 90 °C. Because of that even immersion brewing is selective in a sense, and more often than not this can make for a wonderful cup of coffee! By varying the extraction time and also playing with grind size and brew ratio even more variation becomes possible.
The French press is characterized by the relatively open wire mesh filter. A significant amount of fines will make it into your cup and contribute body and texture to your beverage, and even leave a sediment in the bottom of the cup. Try to touch the sediment and notice how your fingertips can sense the tiny particles. Then consider that our tongue is even more sensitive at detecting particles! The extended extraction of fines that are never separated may also create some undesired bitterness. For a clearer and cleaner taste French press coffee could be filtered through a paper filter. This removes fines and even some of the oils. I would highly recommend that you try this – if only for the sake of the experiment! Nevertheless, immersion coffee can still be a bit muted, even if filtered.
So where does this leave us? Pour over brewing is good, but complex. Immersion brewing is robust, but not perfect. Could it be that Gale Boetticher’s impressive build from Breaking Bad (S3:E6 “Sunset”) is the ultimate way to extract coffee? No, but the clip where he explains his coffee extraction to Walter White really does capture the essence of what we are all longing for: The very best cup of coffee we have ever tasted! In the next part of this series about coffee I will share some novel ideas that hopefully may bring us a bit closer.
Previous posts in the coffee series:
Wonders of extraction: Coffee (part 1)
What’s in a cup of coffee? (part 2)
I just read about walnuts/nocino. I am afraid to make it because of juglone toxicity using wild black walnuts. This site is so cool, and I love the idea of chemistry and food combined. I saw this was your most recent post, maybe you took a hiatus. But I will read previous posts to find out what you and the site are all about. Cheers!
Hi Cris! It’s true that black walnut produces chemicals that are toxic for other plants – this gives the walnut tree a certain advantage. But I would not be worried about the juglone content in a Nocino, but make sure you use gloves – unless you’re OK with stained fingers. Yes – I’m taking a blogging hiatus now. I guess there’ll be another post as we move into winter season.