Flavan-3-ols: the flavanol subclass behind tea leaf and cocoa seed

Two cups, one family of molecules

1. Green tea leaves a dry grip across the tongue and dark chocolate a bitterness that rounds off as it goes, and both come back to a single flavonoid subclass: the flavan-3-ols [1]. Different plants, different processing, same chemistry underneath.
2. Flavanols and flavan-3-ols are two names for one and the same group. It is a subclass of the flavonoids, and its siblings in that family are the anthocyanins, the flavones and the flavonols [1]. Four subclasses, one shared skeleton, four different chemical stories.
3. Catechins are the flavan-3-ols most people associate with green tea leaf. Epicatechin is the name that keeps turning up around cocoa seed. Same subclass, separate plants, and a slightly different set of individual molecules on each list.
4. The flavonoid frame is fifteen carbons: two aromatic rings joined by a bridge of three carbons, and that bridge closes up into the middle ring of the molecule [1]. Everything in the family is a variation on that arrangement.
5. What marks a flavan-3-ol out is an absence rather than an addition. In the middle ring, the feature that other flavonoid subclasses carry is simply not there [1]. One structural detail, one whole subclass.
6. Single units rarely stay single in a plant. Catechin and epicatechin link to one another and build up into proanthocyanidins, the compounds also known as condensed tannins [1]. Small building blocks, considerably larger finished structures.
7. The classic sources read short and familiar. Tea leaf, cocoa seed, grape seed, apple skin. Four tissues from four unrelated plants, each with its own flavan-3-ol profile, and all four of them things people already eat and drink.
8. Flavonols, spelled with an o, are a separate subclass entirely. Quercetin, kaempferol and myricetin sit there [1]. They are not catechins, they do not behave like catechins, and the one-letter gap in the name is doing a lot of work.
9. Hemp is a flavonoid producer, but it is not a flavan-3-ol crop. Its documented profile runs to flavones and flavonols such as apigenin, luteolin, quercetin and kaempferol, plus the cannflavins, prenylated flavones reported in cannabis and, to date, nowhere else [2].

Fifteen carbons, and one feature that isn't there

Every flavonoid starts from the same frame. Fifteen carbons. Two aromatic rings, connected by a bridge of three carbons, and that bridge closes into a third ring that sits in the middle of the molecule [1]. Anthocyanins share that frame. So do flavones, flavonols and flavan-3-ols. The family resemblance is real, and it is the reason all of these compounds get grouped together in the first place.

The differences are local. What hangs off the rings, how many hydroxyl groups there are and where they sit, and above all what is going on in that middle ring. For the flavan-3-ols, the defining point is negative: the feature other flavonoid subclasses carry in the middle ring is absent [1]. That is the marker. Not a decoration added on, but something not present.

The formal name is worth a moment, because it states the structure out loud. A flavan skeleton, with an -ol at position three. Nothing to misread, nothing to guess at. Chemists write flavan-3-ol for exactly that reason, while food and nutrition writing tends to shorten it to flavanol. Both refer to the same set of compounds, and if precision matters, flavan-3-ol is the term that carries the structure with it.

• Tea leaf is the source most closely tied to the catechins, which is why green tea keeps appearing in any discussion of this subclass.
• Cocoa seed is where epicatechin gets named, and the cocoa flavan-3-ol profile is what turns up in chocolate work.
• Grape seed carries flavan-3-ols in the seed tissue rather than the pulp, which is why the seed gets pressed and studied separately.
• Apple skin holds its share too, and it is one reason peeling an apple changes its composition, not just its texture.
• Anthocyanins, flavones and flavonols are the neighbouring subclasses on the same fifteen-carbon frame [1], and none of them are flavan-3-ols.

Units that link into condensed tannins

Flavan-3-ols are joiners. Catechin and epicatechin link to one another, and the result is the proanthocyanidins, a family that also answers to the older name condensed tannins [1]. Two units, three units, considerably more than that. The frame stays the same, but the finished molecule is a great deal bigger than either building block, and it carries far more hydroxyl groups.

That is where the chemistry stops being an abstraction. A long chain of linked flavan-3-ol units has two properties at once: a lot of hydroxyls, and a large surface that can flex and fold. Both matter, because they let the molecule find several points of contact on a protein at the same time. A short chain has fewer options. A long, flexible one has many. Same subclass, quite different behaviour, and the difference comes from the chaining rather than from the individual unit.

Many hydroxyls, one protein

The mouth feel follows directly from that. According to the flavonoid overview published by Panche and colleagues in 2016 [1], the numerous hydroxyl groups and the large flexible surface of these chained structures bind protein and pull it out of solution as a precipitate, which is the mechanism behind astringency. A heavily steeped tea and an underripe apple are the everyday version of the same event [1]. Nothing is being damaged. Protein is simply being bound and dropped out, and the tongue registers that as dryness and grip rather than as a taste in the usual sense [1]. Which is also why the sensation builds the longer a leaf sits in hot water, and why it fades as fruit tissue changes with ripening [1].

What the compound does inside the plant

None of this is built for the sake of flavour. Flavan-3-ols are defence compounds, and their role in defence is what decides where the plant stores them and in which tissues they accumulate [1]. Seed coats, skins, leaves. The outer and the valuable, in other words, rather than evenly spread through the whole organism.

Break that tissue and the picture changes fast. Oxidising enzymes reach the flavan-3-ols and turn them into brown polymers, which seal the wound and leave a chemically hostile surface behind for fungi and bacteria arriving at the break [1]. Cut an apple, bruise a tea leaf, and the darkening you see is that reaction running in real time [1]. It is also the point where the word antioxidant earns its keep in a strict sense: these compounds are reducing agents, chemically described by how they behave towards oxygen [1]. That is a laboratory statement about a molecule, not a claim about a person.

Ripeness, colour and the light on top of a leaf

Flavan-3-ol content in fruit falls as ripening progresses, and the astringency that comes with it works as a timing signal: immature seed tissue reads as rough and unappealing, while a seed that is actually ready sits in softer, less astringent tissue that helps it travel away from the parent plant [1]. A schedule, written in chemistry. The same compounds also do work with colour without having any of their own. Flavan-3-ols are colourless, but through copigmentation they stabilise anthocyanin pigments and shift the hue those pigments show [1]. And in the leaf epidermis they absorb ultraviolet light, contributing to the flavonoid sunscreen layer that sits over the tissue underneath [1].

One letter, a different subclass

Flavanol and flavonol look like typing errors of each other. They are not. Flavonols, with the o, are their own flavonoid subclass, and the names in that column are quercetin, kaempferol and myricetin [1]. Flavan-3-ols are the catechins and epicatechin, plus the proanthocyanidins they build [1]. Two different subclasses of the same family, two different structural patterns, and no overlap in the reference names.

The distinction itself is clean. The spelling is what makes trouble, especially in text that has been through several rounds of editing. This is the practical argument for writing flavan-3-ol whenever accuracy counts: the term states the structure, and it cannot be misread as something else. Say flavanol in conversation about tea and cocoa by all means. Write flavan-3-ol when someone is going to rely on the sentence.

Where hemp sits on this list

Hemp belongs in this conversation for one reason and one reason only: it makes flavonoids, so people reasonably ask whether it makes these. It does not. Hemp is not a flavan-3-ol crop. The flavonoids documented in cannabis are flavones and flavonols, including apigenin, luteolin, quercetin and kaempferol, together with the cannflavins, a set of prenylated flavones reported from this plant and, to date, from no other, as catalogued by ElSohly and Slade in 2005 [2]. Different subclasses, different structures, different plant.

So the honest placement is straightforward. If you want flavan-3-ols, you look at tea leaf, cocoa seed, grape seed and apple skin. If you want the cannabis flavonoid profile, you look at the flavone and flavonol column instead [2]. Working with cannabinoids since 2014 has taught us that naming things accurately removes more uncertainty than stretching a category to fit. Same standard here.