Part 3: Pathways

In contemporary post-harvest processing, fermentation is used to enhance coffee’s quality.  These quality improvements are achieved by allowing the flavours and aroma precursors formed during fermentation to diffuse into the bean.  In part 1 of this article series, we discussed the role of fermentation, and in part 2, we examined the role of diffusion.  However, there is one final aspect to consider: the alternative fates that aroma and flavour precursors may succumb to during fermentation.

In other words the stated intention of enhancing coffee quality through fermentation does not guarantee this realization.  Even if the desired aroma and flavour precursors are successfully produced, there are four potential outcomes only the first of which would be considered optimal [1]:

1. The compound diffuses into the bean and remains unchanged.

2. The compound diffuses into the bean and is metabolised by the bean.

3. The compound does not enter the bean and is metabolised externally.

4. The compound reacts chemically with its surroundings, either internally or externally to the bean.

These parallel pathways have a significant impact on the concentration gradients that guide the diffusion direction of aroma and flavour precursors.  This, in turn, affects the overall quality of coffee by influencing the concentration within the beans.  Apart from the first section (on what occurs under “ideal” conditions) the following discussions are meant to set the stage for future articles that will explore each of these topics in greater detail.

When coffee enthusiasts embark on the journey of fermenting coffee, they likely imagine everything going according to plan.  The microbes working their magic on coffee's mucilage layer, transforming it into aroma and flavour precursors that are then transported into the bean’s depths.  These molecules then patiently await the moment of roasting, when they emerge to augment the delightful bouquet of coffee aromas.  Not entirely untrue.

Between 2020 and 2023 Hadj Salem and associates studied how different molecules behaved during coffee fermentation and they found that several molecules (lactic acid, alanine, glutamic acid and 2-phenylethanol) were consistent with this ideal behaviour, described above, making their influence on coffee’s sensory profile more predictable.  

For instance, lactic acid is not natively found within green coffee but rather is introduced during fermentation.  This means that as it is not degraded by the bean, or the external environment.  Its concentration increases until equilibrium conditions, discussed in the previous article, have been achieved.  The accumulation of lactic acid in the bean will modulate coffee’s sensory profile in two predictable ways (please feel free to share your own experiences with me in the comments below):

• If too much lactic acid is accumulated and survives the roasting process any milk beverage prepared with this coffee will likely taste like the milk is off.  Milk sours when lactose is converted into lactic acid, so the surplus of lactic acid unbalances the flavour profile of the milk triggering a perception that the milk has turned.  This can be considered a “tell” for lactic acid fermentation in coffee.  It is also why I suspect that Nespresso’s La Cumplida Refinada [2] underwent lactic acid fermentation.

• Lactic acid lowers the bean’s pH [3], this influences the Maillard reaction pathways [4] towards those associated with “candied” and “sweet” notes, which coincidently are also descriptors of Nespresso’s La Cumplida Refinada.

Unlike lactic acid, amino acids (alanine and glutamine) are natively present within green coffee beans.  In the app modelling the diffusion of aroma and flavour precursors into the bean these native concentrations are represented by the parameter B.  When these molecules behave in an "ideal" manner we anticipate that they will either remain constant, indicating that the microbial culture does not produce (the) amino acid(s), or they will increase over time when the microbial culture actively produces them.  

The difference between an amino acid producing microbe [5] and one that does not, does not have to be great, as between microbial species, but can be the result of subtle metabolic differences, like those found between strains of the same species.  These nuanced differences may result in significant differences in coffee quality.  For example, if one microbe releases alanine (an amino acid known to be produced by certain microbial strains) while the other does not, and we assume that alanine participates in the Maillard reaction [6] during roasting, then we might anticipate that coffee fermented with the alanine-producing yeast will exhibit stronger fruity (like fresh dates), sweet, and floral notes under acidic conditions (Wong et al., 2008) compared to coffee fermented with the other strain.  Therefore alanine producing yeasts may contribute, along with other yeasts, to the “wild fruity notes” descriptors associated with Nespresso’s La Cumplida Refinada coffee [7].  Other yeasts may digest phenylalanine, by way of the Ehrlich pathway, into 2-phenylethanol (floral, rose-like aroma), another molecule that is accumulated by the green bean.

The difference between dry and wet fermentation will be the involvement of the bean’s own metabolism.

In dry fermentation, as used by Nestlé in La Cumplida Refinada, the husk of the coffee cherry is retained and the microbes are applied on their surface.  Since the husk contains a phytohormone that restrains the beans urge to germinate, the bean is essentially metabolically dormant [8].  This means that we anticipate that the aromas and flavour precursors produced during fermentation and capable of entering the bean will remain constant and not decrease over time.  However, I have yet to come across a study that verifies this occurs in dry fermentation.  It is possible that as the cherry’s flesh breaks down the phytohormone is removed, allowing germination to gradually commence, but this would require a targeted study to confirm.  

A side note for those interested in the fashionable fermentation trends.  A similar “dormant” affect in the bean can be achieved during cryo-maceration [9].  Cryo-maceration is when the beans are frozen before further processing.  This doesn’t actually put the bean to sleep, but rather kills it altogether [10], so the bean’s germination will not contribute to the aroma profile of these coffees because the bean is dead.  Killing the bean is generally seen as detrimental to the green bean’s shelf-life.  The reason behind this will be discussed in a future article on green bean shelf-life.

Conversely, we know that the partial or complete removal of the husk and submergence of the beans in water, during wet processing [11] and fermentation, activates the beans’ germination metabolism.  Even so there has been significant controversy surrounding the degree of influence the bean’s metabolism has over coffee’s quality, especially in fermented coffees where the microbes may be competing for the same nutrients. 

Let’s settle this debate by discussing what happens to the favourite food of both microbe and bean alike, sugar.  Green bean sugar content plays a key role in the development of sweetness and body in the final cup [12].  In a set of experiments conducted in the presence and absence of a microbial competitor, Hadj Salem et al. (2022) established that fructose (simple sugar) was gluttonously consumed by the bean, irrespective of whether the microbe was present.  Although the metabolic rate may vary between varieties, their study clearly conveys that the beans’ sugar content depletes during fermentation, due to the voracious appetite of the bean rather than the microbial presence.  This behaviour suggests that these coffees will have a muted sweetness complimented by a more delicate body [13].  In the app provided the consumption rate of fructose, or another molecule, can be modulating by adjusting the kB parameter.  

Hadj Salem and colleagues (2022) suggest that isoamyl acetate (smells like banana) and butanal (pungent cocoa, malty aroma [14]) may also be metabolised within the bean – something not outside the realm of possibilities [15].  Unfortunately, although there is a noticeable trend of decreasing concentration of these compounds within the green bean during fermentation, the results are not statistically significant, except for butanal under a specific experimental condition that we will explore in the following section.

Fermentation relies on the microbes feeding and reproducing, which are signs of an active metabolism.  Therefore, it's reasonable to assume that these microbes are not only capable of producing aroma and flavour precursors, but also consuming them.  Actually we know this happens.

As the butanal, in Hadj Salem and associates’ (2022) study, shows a significant decrease in the presence of the chosen microbe, a strain of Saccharomyces cerevisiae,  relative to its absence, I suspect this aldehyde is being metabolised, into butanol (not measured).   Yeasts have numerous different pathways, but the last step of the Ehrlich pathway converts aldehydes into alcohols, which is what I expect is happening.  If the compound is indeed sticking to the surfaces of the yeast's cell walls as claimed by the authors, then this effect would be categorized under the following section.

The Ehrlich pathway is known for converting amino acids into their higher alcohol through a variety of steps capable of producing compounds widely praised for their aromatic character, including:

If you've tasted a variety of fermented coffees, you may have noticed that many of the wonderful intense aromas associated with them can be found in this list.  This suggests that this pathway plays a significant role in enhancing the quality of fermented coffees.  This is also why I suspect the green coffee butanal concentrations, in Hadj Salem et al.’s (2022) study, decrease in the presence of yeast.  The yeast consumes the compound, decreasing the external concentration, resulting in an osmotic imbalance causing the compound to diffuse out of the bean, but this is just a hypothesis.  Sensorially, if butanal is transformed into butanol and re-enters the bean, rather than adding a chocolatey aroma, it would instead support a more fruity, fermented character [16].  

While we cannot be sure of butanal’s fate, this example conceptually demonstrates how a microbe’s metabolism, during fermentation, may influence the bean’s composition and potentially the resulting quality.  

A fermentation tank contains thousands of molecules, from the mucilage, to the microbes and beans, down to what’s in the water.  Therefore it is not improbable that there are additional chemical reactions taking place, beyond those being carried out by the living organisms (coffee beans and microbes).  

While this topic is rarely covered in literature, we can safely assume that there are at least a few chemical reactions that could impact the cup’s profile.  For instance, when microbes are used that significantly drop the environment’s pH, it would not be unreasonable to assume acid hydrolysis is facilitating in the digestion of the mucilage.  In essence the acid is facilitating in the degradation of the proteins, carbohydrates and lipids into their smaller components; peptides, amino acids, sugars, glycerol and fatty acids.  As chemical reactions do not discriminate between the structures they degrade they can influence the viability of the organisms within solution (microbes and beans [17]). 

Furthermore, aldehydes such as butanal are also prone to reacting with other molecules (nucleophilic addition, aldol condensation . . . etc.).  To give you an idea of how reactive these aldehydes can be, they are often the active ingredient in spray tanning products that are responsible for your skin turning brown when applied [18].  Therefore, without additional research, we cannot rule out the possibility that instead of being transformed into a different aroma compound by the yeast, butanal might simply be interacting with one of the many other molecules, potentially losing its sensory relevance and being swept away with the wastewater.

If dead or dormant coffees serve as good carriers of the fermented aromas and flavour precursors, why aren't we using this technology to revive "past crop" coffees?  "Past crop" coffees are aged green coffees that have been stored for a year or more and generally have lost their viability, i.e. they are the aroma skeletons of the coffee world.  These coffees decrease in value as they lose their initial youthful character and develop a paper/cardboard [19] taste when roasted, which many consumers, including myself, find unpalatable.  Since only the minority of the flavours and aromas generated during fermentation are absorbed by the initial (fresh) batch of beans, why not use the residual microbial brew to enhance these older coffees, potentially increasing their market value?  This would allow farmers to earn a better income from a depreciating investment.

Ameliorating “past crop” coffees maybe perceived as, “putting lipstick on a pig” by some, but is it really?  There are some that claim that fermenting coffees increases the green coffee shelf-life by decreasing the bean’s pH, but the bean’s pH naturally decreases during ageing, so wouldn’t that suggest that fermentation facilitates ageing?  Those that believe this are applying broadly known food preservation principles to coffee, but are not taking into consideration the chemical context of what the ageing process entails in green coffee.  The truth is that, the increase in acidity likely facilitates in controlling the microbial growth during the drying process, immediately after fermentation, but would become secondary to the green bean’s water activity, ~0.4, once the coffee had been properly dried, i.e. at the beginning of its shelf-life. 

What is more likely happening, and would be perceived as a longer shelf-life, is that the aromatic microbial metabolites are camouflaging, dare I say “masking”, the natural deterioration of the bean.  One could verify this by examining the viability of the beans.  If fermentation truly preserved the integrity of the bean then an aged fermented bean should sprout equally well when compared to a younger bean of a comparable quality.

This hypothesis suggests that fermenting the beans freshly after harvesting actually adds aromas and flavour precursors that obscures the bean’s natural ageing process.  In other words if the pig already has lipstick on, does it really matter when it’s applied?  What do you think?  Should fermentation be reserved for newly harvested coffees?  Or do you think this process should be extended to giving “past crop” coffees an aromatic facelift?

Naturally there may be challenges, for example dealing with any rancidity within the beans, but these traits may be (already) concealed by the sugars and other strong aromas contributed by the microbial metabolites [20].  If you've experimented with this, please share your thoughts in the comments below.  I'd love to hear about your experiences! [21]

Cup quality is not solely dependent upon the generation of aroma and flavour precursors during fermentation, or their successful diffusion into the bean, but also on whether or not they feed into an alternative pathway.  

Optimally, the aroma and flavour precursor formed diffuse into the bean and accumulate, representing the molecules final destination, before roasting, but this is not always the case.  Some molecules will,

• Be consumed by the bean’s metabolism

• Be consumed by the microbes

• Chemically react with their surroundings

All of these alternative outcomes have the potential to significantly influence the final cup quality (and most of them represent a loss of opportunity for the molecules, which we want to contribute to the brewed coffee).  By understanding how each pathway influences cup quality, we can better orchestrate the desired outcome, and perhaps even turn unwanted pathways into new opportunities.