Winemaking Laboratory
Grape chemistry and harvest / Phenolic compounds

Chapter II · Grape chemistry and harvest

Phenolic compounds

How to assess the phenolic maturity of grapes with the Glories method, how anthocyanins and tannins pass into wine under European, Imeretian, Kartlian and Kakhetian winemaking, and what a wine's colour spectrum tells us.

Glories analysis

Anthocyanins (mg/L) = A520 × 22.76 × dilution. EA% = (ApH1 − ApH3.2) / ApH1 × 100; skin-tannin contribution to A280 = ApH3.2 × 40 / 1000; Mp% = (A280 − skin contribution) / A280 × 100. The interpretation thresholds are guides — they vary by variety and vintage.

Anthocyanin potential ApH1—mg/L
Extractable ApH3.2—mg/L
Extractability EA%—%
Phenolic richness A280—index
Skin-tannin contribution—index
Seed-tannin share Mp%—%

Winemaking method

Illustrative kinetic model: first-order extraction, rates scaled with Q₁₀ = 2; seed-tannin extraction speeds up as alcohol accumulates; anthocyanins gradually turn into polymeric pigments or are lost. Potentials refer to 100% pomace; the default anthocyanin value is ApH3.2 from the first mode.

Monomeric anthocyanins—mg/L
Polymeric pigments—mg/L
Total tannins—g/L
Seed-tannin share—%

Spectrum

Colour intensity and hue are converted to a 10 mm cuvette. The share of the flavylium (red) form is computed for malvidin-3-glucoside: hydration pK = 2.6, quinoidal-base pKa = 4.25.

Colour intensity—10 mm
Hue A420/A520—
Brilliance dA%—%
Red form at this pH—%

Intuition

Sugar and acids are the "watery" part of must — they are almost entirely in the juice. Phenolic compounds, by contrast, sit mainly in the solid parts: the skins, seeds and stems. They are what give wine its colour, body, astringency and ability to age. A wine's phenolic composition therefore depends on two things: how many phenolics are in the grapes, and how many we extract during winemaking.

The two most important groups of phenolics are the anthocyanins — red and violet pigments found only in the skins of black grapes (and, in Saperavi, in the flesh too) — and the tannins, which bind proteins and produce the drying, "gripping" sensation in the mouth. Tannins are found in the skins, the seeds and the stems, but they differ in quality: skin tannins are soft, seed tannins coarse and bitter, especially in unripe seeds.

This topic matters especially for Georgian wine: in the Kakhetian method white grapes stay on the pomace for months, so an "amber" qvevri wine can be ten times richer in phenolics than a European white wine. On this page we learn how to assess phenolic maturity, how extraction proceeds over time and how to read a wine's colour.

History

Scientific study of wine colour and astringency began in the mid-20th century. In 1958 the French researcher P. Sudraud introduced colour intensity (the sum of absorbance at A420 and A520) and hue (their ratio) — the two numbers we still use to describe wine colour. In 1970 Pascal Ribéreau-Gayon used absorbance at 280 nm to estimate total phenolics, and in 1984 Yves Glories added a blue component (A620) to colour intensity.

In 1971 the Australian chemist T. Somers showed that the colour of old red wine comes not from monomeric anthocyanins but from polymeric pigments — compounds of anthocyanins and tannins. In 1977 R. Brouillard and J.-E. Dubois described the equilibrium of anthocyanin forms as a function of pH, which explained why high-pH wine loses colour.

In the 1990s at the University of Bordeaux, Y. Glories and co-workers developed a method for assessing phenolic maturity: grapes are extracted in two parallel solutions, at pH 1 and pH 3.2, and comparing the two results tells us how much colour is in the berry and how much can be extracted under wine conditions. The method is still the reference standard for phenolic maturity.

Phenolic compounds in grapes

Phenolics are compounds with an aromatic ring carrying one or more hydroxyl groups. Grapes contain two large classes. Non-flavonoids (hydroxycinnamic acids, resveratrol) are mainly in the flesh and juice and make up a large part of the phenolics of white wine. Flavonoids — anthocyanins, flavonols and flavan-3-ols (catechins and their polymers, the condensed tannins) — are in the skins, seeds and stems.

Anthocyanins begin to accumulate at véraison and reach a maximum at ripeness; in a hot climate (above 35 °C) their synthesis stalls and part of them breaks down. The main anthocyanin of Vitis vinifera is malvidin-3-glucoside. Saperavi is a rare example among Georgian varieties of a grape with coloured flesh (a so-called "teinturier"), so its anthocyanin potential is very high.

The main difference between tannins is in quality. Skin tannins are long, less galloylated polymers with soft astringency. Seed tannins are short and more galloylated — coarse and bitter. As they ripen, seeds turn from green to brown, their tannins "lock up" through oxidation and their extractability falls. Stem tannins are coarse and give green, herbaceous notes if the stems are unripe.

Mathematics of the Glories method

Anthocyanin concentration is estimated from A520 (520 nm, green light that the red pigment absorbs). At pH 1 all anthocyanins are in the red flavylium form, so absorbance reflects their total amount:

$$A_{pH} = A_{520} \cdot 22.76 \cdot D\quad (\text{mg/L})$$

where 22.76 is a factor derived from the absorption coefficient of malvidin-3-glucoside and $D$ is the dilution. The pH 1 solution fully breaks down the cell walls and gives the total anthocyanin potential ($A_{pH1}$); the pH 3.2 solution mimics wine conditions and gives the extractable part ($A_{pH3.2}$). The extractability index:

$$EA\% = \frac{A_{pH1} - A_{pH3.2}}{A_{pH1}}\cdot 100$$

The lower EA%, the riper the skin cell walls and the more easily colour passes into the wine. Typical values are 20–70%.

Phenolic richness is estimated from the absorbance of the pH 3.2 extract at 280 nm (1:100 dilution): $A_{280} \cdot 100$. Within this value the skin-tannin contribution is estimated in proportion to anthocyanins — $d_{\text{skin}} = A_{pH3.2}\cdot 40/1000$ — and the rest is attributed to the seeds:

$$Mp\% = \frac{A_{280} - d_{\text{skin}}}{A_{280}}\cdot 100$$

A high Mp% means a large share of the phenolics comes from the seeds — the seeds are still green and their tannins extract easily. Mp% falls as the grapes ripen.

Extraction kinetics

During maceration (time on the pomace) each component passes into the wine at its own rate. The calculator describes this with first-order kinetics: $dC/dt = k\,(C_\infty - C)$, where $C_\infty$ is the extractable potential and $k$ the rate constant. Temperature acts through $Q_{10} = 2$: $k(T) = k_{25}\cdot 2^{(T-25)/10}$.

Anthocyanins extract quickly even in aqueous solution ($k \approx 0.45$ day⁻¹), but they are lost at the same time — adsorbed onto yeast and pomace surfaces, broken down, and combined with tannins into polymeric pigments. Their concentration therefore peaks within 4–7 days and then falls:

$$A(t) = A_\infty\,\frac{k_1}{k_1 - k_2}\left(e^{-k_2 t} - e^{-k_1 t}\right)$$

Skin tannins are also extracted early ($k \approx 0.2$ day⁻¹). Seed tannins are protected by a waxy coat and dissolve poorly without alcohol, so their extraction rate rises as fermentation proceeds: in the model $k_{\text{seed}} = 0.005 + 0.08\,e^{1.5}$, where $e$ is the degree of fermentation (0 to 1). Stem tannins pass over slowly but steadily ($k \approx 0.05$ day⁻¹). During long maceration a small part of the tannins is lost with proteins and lees.

This model is illustrative: real rate constants depend on variety, ripeness, degree of crushing and pomace management (punching down, stirring). Its value lies in its shape — it shows why skins dominate in short maceration, and seeds and stems in long maceration.

Colour and pH

An anthocyanin exists in water in several forms. The red flavylium cation ($AH^+$) takes up water and becomes a colourless hemiketal (pK ≈ 2.6), and partly loses a proton to form a violet quinoidal base (pKa ≈ 4.25). The share of the red form:

$$f_{AH^+} = \frac{1}{1 + 10^{pH - 2.6} + 10^{pH - 4.25}}$$

At pH 3.2 this is about 20%, at pH 3.8 only 6%. In other words, most anthocyanins in wine are colourless, and even a small rise in pH visibly reduces colour. Polymeric pigments and copigmentation (anthocyanins "stacking" with other phenolics) weaken this effect — which is why the colour of aged wine is more stable to pH and SO₂.

Spectral colour characteristics (for a 10 mm cuvette): colour intensity $CI = A_{420} + A_{520} + A_{620}$; hue $A_{420}/A_{520}$; red brilliance $dA\% = \left(A_{520} - \frac{A_{420}+A_{620}}{2}\right)/A_{520}\cdot 100$. The hue of a young red wine is 0.4–0.6 and rises to 0.8–1.2 with age — the colour shifts toward brick.

Key quantities

QuantitySymbolUnitTypical value
Anthocyanin potentialApH1mg/L800 – 2500 (Saperavi higher)
Extractable anthocyaninsApH3.2mg/L400 – 1200
ExtractabilityEA%%20 – 70
Phenolic richnessA280index30 – 80
Seed-tannin shareMp%%20 – 70
Tannins in red wine—g/L1 – 4
Phenolics in European white wine—g/L0.2 – 0.4
Phenolics in Kakhetian white wine—g/L1.5 – 3
Colour intensity (red)CI—5 – 25
HueA420/A520—0.4 – 1.2

Extraction model parameters (25 °C)

ProcessRate k (day⁻¹)Note
Anthocyanin extraction0.45fast even in aqueous solution
Anthocyanin loss0.035half becomes polymeric pigments
Skin tannins0.20early, during fermentation
Seed tannins0.005 – 0.085increases as alcohol accumulates
Stem tannins0.05slow, steady
Tannin loss0.003with proteins and lees

Connections

On the "Ripening dynamics" page of Chapter I, phenolic maturity (anthocyanins and seed maturity) is computed alongside technological maturity; here the same maturity is assessed with a laboratory method. The previous page, "Acids, pH and the buffer system", computes pH — and that pH determines the share of the red form of colour. The next page, "Harvest and must yield", shows how much pomace is obtained and how it is distributed among qvevri.

In Chapter III (Alcoholic fermentation) we will see how fermentation rate and temperature govern seed-tannin extraction. Chapter IV (Wine chemistry and stabilisation) covers the oxidation and polymerisation of phenolics, and Chapter V the management of pomace in the qvevri — duration of contact, punching down and racking.

Application — Georgian practice

The three traditional Georgian methods differ from one another, and from the European method, precisely in phenolic extraction. In the Kakhetian method the must ferments with the whole pomace, often including stems, and white wine stays on the pomace from autumn to spring; red wine usually stays on it for about a month. In the Imeretian method only a small part of the pomace, about one tenth, is added to the qvevri, without stems. The Kartlian method is intermediate — about one third, with stems. For comparison, in the European method white wine ferments without pomace and red wine stays on it for 1–2 weeks.

As a result, a Kakhetian Rkatsiteli or Mtsvane is far richer in phenolics than a European wine from the same grapes, and its stability — against oxidation and microbes — rests largely on its tannins. But this requires ripe seeds and stems: grapes with green seeds give a coarse, bitter wine under long maceration. That is why assessing phenolic maturity (by the Glories method or simply by seed colour) is especially important for grapes destined for the Kakhetian method.

Glossary

TermDefinition
Phenolic compounds (polyphenols)Aromatic compounds with hydroxyl groups — the source of wine's colour, astringency and body
AnthocyaninsRed and violet pigments in the skins of black grapes
TanninsProtein-binding polyphenols, the source of astringency
Phenolic maturityThe optimum of the amount and extractability of skin and seed phenolics
ExtractabilityHow easily phenolics pass into the wine
Maceration (time on the pomace)Contact of wine or must with the pomace for extraction
Flavylium cationThe red form of anthocyanin, dominant at low pH
Polymeric pigmentsStable compounds of anthocyanins and tannins
CopigmentationEnhancement of anthocyanin colour by association with other phenolics
Colour intensitySum of absorbance at 420, 520 and 620 nm
HueThe A420/A520 ratio — the balance of yellow and red components
Amber wineWine made from white grapes with long contact on the pomace

Practical use in calculations

A Glories analysis was run on a sample taken from a Saperavi plot one week before harvest: the pH 1 extract (dilution 20) gave A520 = 2.95, the pH 3.2 extract 1.60, and A280 was 0.58. The calculator gives an anthocyanin potential of 1343 mg/L and an extractable part of 728 mg/L. EA% = 45.8, i.e. medium extractability. Phenolic richness is 58, of which the skin contribution is 29.1, and Mp% = 49.8 — the seeds are partly ripe. The winemaker decides to make the wine by the Kakhetian method with one month of maceration. The extraction mode (fermentation at 25 °C, then 13 °C) shows anthocyanins peaking at about 590 mg/L on day 6. After 30 days monomeric anthocyanins are about 430 mg/L and polymeric pigments 150 mg/L. Total tannins are about 3.1 g/L, more than half of it from the seeds. The same grapes made the European way (10 days, no stems) would give about 2.0 g/L of tannin. For comparison, white Rkatsiteli by the Kakhetian method (5 months) gives about 2.6 g/L of tannin, by the Imeretian method 0.2 g/L. Finally, the spectrum of the young wine (1 mm cuvette: A420 = 0.45, A520 = 0.95, A620 = 0.18) gives a colour intensity of 15.8 and a hue of 0.47 — a young, purple-red wine. At pH 3.5 only 11% of the anthocyanins are in the red form.

For beginners — from sample to maceration plan

1. Open the "Calculator", mode "Phenolic maturity". Enter the laboratory results: A520 of the pH 1 and pH 3.2 extracts, the dilution and A280. The cards show the anthocyanin potential, extractability (EA%) and the seed-tannin share (Mp%).

2. The point for your sample on the map shows how ready the grapes are. The green zone (low EA% and Mp%) means colour extracts easily and the seeds are ripe — long maceration is possible. The red zone means a short maceration is better, or the harvest should wait.

3. Switch to the "Extraction kinetics" mode. Choose the grape colour and the winemaking method — European, Imeretian, Kartlian or Kakhetian. The anthocyanin potential is taken automatically from the first mode.

4. The chart shows how skin, seed and stem tannins pass into the wine and how anthocyanins change. The lower chart compares all four methods on the same grapes. Try changing the days of maceration and the temperature.

5. The "Wine colour" mode assesses the spectrum of a young or aged wine. Enter A420, A520, A620 and the cuvette path length — you get colour intensity, hue, and what share of the anthocyanins is red at your wine's pH.

For professionals — limits of the method

1. The result of the Glories method depends heavily on sample preparation: grinding, extraction time (4 hours) and temperature must be the same for every sample. Absolute values compare poorly between laboratories — the most valuable information is the trend for one plot, week to week.

2. The EA% and Mp% thresholds (35/55% and 40/60%) are guides. For Saperavi and other high-anthocyanin varieties EA% is often high not because colour is lacking, but because only part of a very large potential can pass into the wine. The absolute value of ApH3.2 is therefore just as important as EA%.

3. Mp% is an indirect estimate: skin tannins are estimated in proportion to anthocyanins (×40/1000), which does not work for white grapes. For white grapes it is better to assess seed maturity directly — by colour (green → brown), firmness and taste.

4. The extraction model parameters are illustrative. If you measure the wine several times during maceration in your cellar (anthocyanins, total phenolics), you can tune the potentials so that the curve matches your data — and then use the model to estimate the outcome of other durations and temperatures.

5. When measuring colour, young red wine is read in a 1 mm cuvette (or diluted) so that absorbance stays within 0.1–1.0. The sample must be filtered — turbidity raises absorbance at every wavelength and artificially inflates colour intensity. SO₂ bleaches anthocyanins, so readings taken before and after sulfiting are not comparable.

1. In a Glories analysis ApH1 = 1400 mg/L and ApH3.2 = 700 mg/L. What does EA% = 50 mean?

EA% shows what share of the anthocyanins remains in the skins at pH 3.2. The lower EA%, the more easily the cell wall gives way and the more colour passes into the wine.

2. Why does seed-tannin extraction increase in the second half of fermentation?

Seed tannins lie under a lipid-waxy coat, and alcohol speeds their extraction. That is why in long maceration (the Kakhetian method) the seed contribution dominates, while in short maceration the skins dominate.

3. Why does red wine lose colour as pH rises?

The flavylium cation hydrates (pK ≈ 2.6). At pH 3.2 the red form is about 20%, at pH 3.8 about 6%. That is why high-pH wine is paler and its colour less stable.

4. After 5 months of Kakhetian maceration there are few monomeric anthocyanins, yet the wine is still dark. Why?

Over time anthocyanins react with tannins and acetaldehyde to form polymeric pigments, which are far more stable to pH and SO₂. They provide most of the colour of aged red wine.