Dea Bardhoshi

Dea Bardhoshi

Data Scientist (UC Berkeley '23)

A decaf project

As a coffee lover who wants to also enjoy coffee in the evenings without risking staying up all night, I like trying out decaf. At home, I mostly make my coffee with a pour over or a French press, and the decaf versions of those coffee drinks come out tasting much different than regular coffee. Trying out these different batches of beans, I realized I don’t really know much about how decaf is made and why it tastes different from regular caffeine coffee. Therefore, this is my attempt at answering some of my own questions.

Preparation: how is decaf made?

Decaf coffee is created by extracting the caffeine from coffee beans using water and various solvents or filters, depending on the specific method. One method you might have seen mentioned in a bag of decaf beans is the ‘Swiss Water process,’ which, I learned, is apparently the name of the company that does it as well! Swiss Water Process collaborates with roasters to source decaf for them, prepared through their special method out of their five factories around the world.

So the details: the Swiss Water Process uses a mixture of water, temperature control, and a proprietary carbon filtration system to remove caffeine without the use of chemical solvents. It starts with a batch of green coffee beans that are soaked in hot water, dissolving their caffeine and flavor compounds. This water, now saturated with all the soluble components of coffee, is passed through an activated carbon filter designed to trap only the caffeine molecules. The remaining liquid, known as Green Coffee Extract (GCE), is then used to soak a new batch of beans, allowing them to shed their caffeine without losing flavor. Since the GCE is already rich in the desirable compounds that give coffee its taste, the beans retain their original character while becoming nearly caffeine-free. This process is repeated until the beans are 99.9% caffeine-free, at which point they are dried and shipped off to roasters.

Out of curiosity, I checked the batch I got from Wimp Decaf a while ago:

Image by Author

Wimp Decaf is a small business out of North Carolina that sells various decaf coffee beans. For a few of their current batches they use a process called Thermal Shock, invented by Wilton Benitez. Unlike Swiss Water, which relies on gradual extraction, Thermal Shock involves rapidly cooling the beans after an initial high-temperature soak, which causes the cell walls to contract and release caffeine more effectively. The process is designed to preserve as much of the coffee’s natural sweetness and acidity as possible.

Their other current coffee offerings use a mix of Swiss Water Process and EA (ethyl acetate) washing. EA is yet another decaffeinating process that is very common, particularly in Latin America. Also known as the ‘sugarcane process,’ EA decaffeination uses ethyl acetate, a compound that naturally occurs in fruits and sugarcane, to bond with caffeine molecules and remove them from the beans. The beans are steamed to open their pores, then soaked in an EA solution, where the caffeine dissolves. After several washes, the beans are rinsed, dried, and sent to roasters. Since EA is derived from sugarcane, it tends to impart a mild sweetness to the beans, making it a popular choice for decaf drinkers looking for a little extra complexity in their cup.

Yet another process, much less common compared to Swiss Water and EA, is carbon dioxide decaffeination. Carbon dioxide (CO₂) decaffeination uses pressurized liquid CO₂ to extract caffeine while preserving the coffee’s flavors. The beans are soaked in water and placed in a sealed chamber, where CO₂ is applied at high pressure, binding to the caffeine molecules and drawing them out. Since CO₂ targets caffeine specifically, this method retains more of the coffee’s original oils and aromas compared to some solvent-based processes.

Different batches and some brewing advice

For the purposes of this blog, I’ve recently used beans from the following roasters and batches (although not directly comparing them). They used a mix of washing methods, which is useful for comparing any taste differences. I mostly prefer light roasted coffee, which is also reflected in my choices below since most batches lean toward light roasts, fruity notes with a lot of acidity. However, I’ve found I had the tendency to amplify the taste of these decafs as I was brewing them and often the coffee came out tasting extra sour (like a really acid fruit). Darker decaf beans likely taste much different, so it’s possible that regular coffee preferences affect decaf tasting ones as well.

The coffees:

Image by Author

This last batch, sourced by Samuel Diego Bermudez in El Paraiso in Colombia, was highly recommended online. Their process uses a mixture of EA and thermal shock which lock in the complex flavors of the coffee. My first thoughts when I got this coffee were that it smelled so good even through the sealed bag. When I got to brewing it, very surprisingly, I could also smell all the fruity notes in the grounds and then taste them in the liquid while drinking. This coffee didn’t taste muted or flat: it tasted like a whole new level of decaf I didn’t know was possible. Also, look how pretty it is! 😍

Here are some guides I’ve found about the differences between brewing decaf and regular coffee, with some handy tips. In general, a lot of advice comes through practical experimentation and overall experience, so in the sprit of that, I decided to do some experimentation of my own.

For instance, I learned that chemically the decaffeination process makes beans more brittle and porous so they release flavors quicker. Therefore, when brewing you should try to avoid overextracting the beans. In practice, for different brewing methods this means changing your technique a little bit. For instance, French press decaf should be brewed for less time than regular coffee (around 3 minutes). On the other hand, if you are making decaf espresso, you should be aiming for coarser grounds and a longer shot time.

Making a pour over

Starting Recipe: temperature 190-210F (90-100C), grind size: coarser than other regular pour over, coffee to water ratio: start with 1:16 and adjust.

My favorite method of making coffee at home is the pour over, so making decaf taste good with this method was one of my biggest goals for this project. While I’m not quite there yet, there were a few factors that were most important for good-tasting decaf: the beans you use, the grind setting and the coffee to water ratio.

What I’d recommend (and what I did) is trying a few different grind settings to see how they taste. Since my current grinder is a handheld one with an adjustment wheel, I can adjust grind sizes approximately but not super precisely. The grinder itself also produces grounds that are not very uniform, so I’ll usually get some finer and some coarser grounds for any given setting. For the purposes of an initial trial, I tried out multiple very different grind settings, one around my usual pour over setting and one much coarser. If you can’t measure the actual diameter, try going for something around the consistency of sea salt.

Playing around a bit with coarser grinds that my usual pour over setting, eventually I got a cup which tasted kinda good and the decaf notes weren’t extremely powerful. However, another issue to keep in mind is that decaf beans themselves will grind inconsistently even with precise grinders due to their brittleness. In these cases, you can try mixing the slurry between pours or even order your coffee pre-ground to your desired prep method.

Another important variable is the coffee-to-water ratio, i.e. the grams of coffee per gram of water. The recommended decaf ratio is around 1:16 (or 30 grams coffee per 480 ml water), which is a starting point for regular coffee as well. Depending on taste preferences, you can adjust this for stronger (less water and more coffee, e.g. 1:15) or weaker (e.g. 1:17) tasting decaf.

More chemistry differences & other decaf follow ups

But, why does decaf really taste different, even after optimizing all these steps (assuming I am not messing things up)? Chemical tests have shown some differences in the specific compounds that give coffee its flavor and smell, which are mainly responsible for decaf tasting so different. For instance, pyrazines are compounds created when coffee is roasted and they give coffee a nutty flavor — it turns out that pyrazine content is much lower in decaf coffee (a study even suggesting 58% lower). Each decaffeination method likely preserves pyrazines in different amounts, however I can’t find any studies comparing between them in decaf coffee. This information is probably the (my?) next step for understanding the different types of decafs out there.

However, some of the more innovative coffees I had showed me that decaf can taste delicious. In particular, the Hydrangea batch but even the CO2 bag, which smelled pretty good and tasted decent. New methods like CO2 decaffeination, thermal shock, and fermentation are all very exciting directions for the world of decaf coffee. In any case, while I’ve learned a good amount from this project, there is still much more decaf that I’d like to try in the future. So I made a spreadsheet of batches people have recommended I plan to drink as well. You can check it out here: Decaf Spreadsheet.

☕️ -Dea