Skip to content
CiderHQ
Search

Compound

Procyanidins

The condensed tannins of cider fruit, whose chain length — not their quantity — decides whether the mouth registers bitterness or astringency.

Also called Condensed tannins, Proanthocyanidins, Apple tannin.

Class
Phenolics
Formula
Not a single molecule — see below
How often it matters
Present in every cider

What it does in cider

How it is perceived

What the compound registers as, and at roughly what concentration. Perception is not a property of the molecule alone: sugar, tannin and carbonation all change where a threshold falls.

On threshold

There is no single threshold, because two effects rise on different curves. Bitterness peaks with the smaller oligomers and falls as the polymer lengthens; astringency does the opposite, becoming perceptible only once chains are long enough to cross-link salivary proteins. Astringency also accumulates over successive sips instead of adapting away, so a cider that seems moderate on first taste can be markedly drying by the third.

Descriptors it is responsible for

Sensory records that name Procyanidins as a cause. Each states the perception and the mechanism behind it.

The structure it moves

Structural dimensions this compound contributes to. Direction and amount depend on concentration and on what else is in the drink; the dimensions themselves are set out in full under Sensory.
DimensionWhat it is
BitternessA taste sensed at the back of the tongue, distinct from the drying grip of astringency.
AstringencyThe drying, rough sensation left after swallowing.
TanninThe phenolic material that gives cider structure, grip and ageing capacity.
BodyHow much weight and viscosity the drink has in the mouth.
Phenolic characterSmoky, spicy, leathery or medicinal aromas that sit apart from fruit.

Measured figures

Shown as they were measured, with the context each was taken in. They are not averaged: a concentration recorded in one country's fruit in one decade is not a constant.

3 separate analyses of total tannin. They are shown as they were measured, in their own contexts, and are not averaged — the same fruit grown somewhere else can genuinely give a different number.

Total tannin0.050–0.600 % w/v

Long Ashton, Somerset, England, 1903–2003 · National Fruit and Cider Institute / University of Bristol

Historic juice analyses spanning dessert fruit at the bottom of the band to the most tannic bittersweets at the top. The figure is total tannin by a colorimetric method: it says nothing about how that tannin is distributed between short and long chains, and that distribution is what decides how the cider tastes.

Total tannin0.200 % w/v

context not recorded · Long Ashton classification threshold · National Fruit and Cider Institute / University of Bristol

The line that separates bittersweet and bittersharp fruit from sweet and sharp fruit under the English scheme. A convention chosen to be useful for blending, not a point at which anything changes chemically.

Total tannin2.4 g/L

Normandy and Brittany, France · Thiolysis and HPLC of centrifuged juices from French cider apple varieties · Guyot, Marnet, Sanoner and Drilleau, Journal of Agricultural and Food Chemistry 51(21):6240–6247

The upper figure the study reached, in centrifuged juice — that is, after the solids that carry much of the phenolic load have been taken out. It is a ceiling observed in a variety trial, not a typical juice: most of the fruit measured sat well below it, and variety was the largest single source of variation.

The phenolic fraction responsible for bitterness and astringency. Measured in percent weight per volume.

Procyanidin chain length25.00 :1

Normandy and Brittany, France · Average degree of polymerisation by thiolysis, one variety, centrifuged juice · Guyot, Marnet, Sanoner and Drilleau, Journal of Agricultural and Food Chemistry 51(21):6240–6247

A twenty-five-unit average chain, dissolved in the juice at close to 1.2 g/L. This matters because the intuition that big polymers fall out of solution is wrong here: they stay in, and they arrive on the palate. The same study found that the average chain length in juice is significantly shorter than in the fruit it came from, so pressing itself shifts the distribution.

The average degree of polymerisation, which decides how much of the phenolic load reads as bitterness and how much as astringency. Measured in ratio.

What forms it

Processes that put this compound into the drink, or increase how much of it is there.

What removes or limits it

Processes that reduce it, hold it below a threshold, or stop it forming in the first place.

Faults it is implicated in

Being implicated is not the same as being a fault. Several of the compounds on this site are ordinary constituents of a sound cider and define a named fault only above a concentration.

Chain length splits bitterness from astringencyAt a fixed 750 milligrams per litre in model cider, astringency rises with procyanidin chain length while bitterness peaks in the middle and falls away.Astringency — rises with chain lengthBitterness — peaks at the tetramerMonomerDimerTetramerLong polymerAverage degree of polymerisation →Held constant: 750 mg/LConcentration, ethanol, fructose andmalic acid identical at every point.
At a fixed 750 milligrams per litre in model cider, astringency rises with procyanidin chain length while bitterness peaks in the middle and falls away.
Described in full
Shape
A horizontal axis of increasing average degree of polymerisation, marked with four positions: monomer and dimer at the left, tetramer in the middle, and long polymer at the right. Two labelled bands run across it — an astringency band that thickens steadily from left to right, and a bitterness band that is thin at the left, thickest at the tetramer, and thin again at the right.
Astringency
Increases with the degree of polymerisation. Longer chains carry more sites to cross-link salivary proteins, and it is the loss of lubrication in the mouth that reads as drying rather than as taste.
Bitterness
Peaks at the medium chain length. In the study behind this figure, the tetramer fraction was the most bitter of the four tested; both the shorter and the longer fractions were less so. Bitterness is a receptor event, and a molecule can be too large to fit the receptor while remaining perfectly able to bind a protein.
The fixed condition
Every point on this axis is 750 milligrams of procyanidin per litre in the same model solution of water, ethanol, fructose and malic acid. Concentration is held constant so that the axis shows chain length alone. The companion study found the chain-length effect is not identical at every concentration, so this figure describes one concentration.
The side effect
A note beneath the axis records that in the same experiments the perception of sweetness and sourness also moved with procyanidin concentration, although the fructose and the malic acid were identical in every sample. Tannin does not only add its own sensations; it changes the reading of the others.
What this is not
A model solution is not cider. It contains no pectin, no protein, no carbonation and none of the hundreds of compounds a fermentation makes, and every one of those can shift how a phenolic behaves on the palate.

About Procyanidins

A procyanidin is a chain of flavan-3-ol units joined end to end. In apples those units are overwhelmingly (−)-epicatechin, with (+)-catechin appearing mostly at the end of a chain, and the links are carbon–carbon bonds that acid and enzymes do not readily break. Chains of two, three and four units are common; so are chains of twenty and more. The whole population in a juice is usually summarised by its mean degree of polymerisation, and that single number explains more about how a cider tastes than the total tannin figure ever will.

The reason is that bitterness and astringency are different phenomena with different size requirements. Bitterness is a taste: a molecule has to fit a receptor on the tongue, and only the smaller oligomers are small enough to do it well. Astringency is not a taste at all. It is what happens when a long, flexible polymer with many hydroxyl groups cross-links the proline-rich proteins in saliva and precipitates them, stripping away the film that normally lets the mouth slide against itself. The tongue reports roughness and drag. Short chains cannot cross-link enough protein to do this; long chains can, and the longer they are the better they do it, up to the point where they become too insoluble to stay in the drink at all.

So two ciders with identical total tannin can be entirely different propositions. One built on short oligomers is bitter and comparatively soft; one built on long polymers is drying and comparatively unbitter. This is why the Long Ashton classification, which measures total tannin, explicitly does not predict the balance of bitterness and astringency, and why cultivars with similar analytical figures have such different reputations among makers. It is also why the mouth-feel of a cider changes with time: procyanidins go on polymerising in the tank and the bottle, so bitterness recedes, astringency first rises and then falls as the longest chains drop out as sediment.

Everything a maker does between the mill and the bottle moves this. Procyanidins sit in the skin and flesh and, in high concentration, the seeds, so how finely the fruit is milled and how long the pomace macerates decides how much is extracted. Polyphenol oxidase attacks them in the presence of air, coupling them into larger, browner and less soluble structures — which is exactly why deliberately oxidised juice makes a softer, less bitter cider than sulphited juice from the same fruit. Protein fining works by mimicking saliva: gelatin or isinglass added to the tank precipitates the longest chains preferentially, taking astringency out and leaving bitterness comparatively untouched. And in the glass, the same protein binding happens with food, which is why a tannic cider reads quite differently alongside a hard cheese or a fatty pork dish than it does alone.

Related compounds

Compounds it is formed from, converted into, confused with, or routinely met alongside.

What people ask next

Questions readers ask about the things this page mentions. Each one goes to the section that answers it rather than to a page written to receive the question.

Where to go next

Sources

What this page rests on. Where a source is marked as registered rather than read, CiderHQ is recording that the body is authoritative on the subject without claiming to have worked through the document itself. See our evidence policy for what each state means.