Chapter IV · Stabilisation and clarification
How much oxygen wine takes up during racking, pumping, filtration and bottling, how it uses it up, when oxygen helps and when it ruins the wine; and the opposite problem: where rotten-egg and cabbage smells come from and how to remove them with copper.
Typical figures (mg/L O₂ per operation): open racking 3–6, siphon 0.5–1.5, pump 1–2, filtration 0.5–2, topping 0.2–0.5, bottling 1–4 (at home, without inert gas). "Capacity" is how much oxygen a wine tolerates without significant oxidation: white ≈ 10–20, amber ≈ 30–50, red ≈ 40–80 mg/L (approximate; depends on phenolics and SO₂).
Dissolved oxygen is consumed with first-order kinetics: k = k₂₀·2^((T−20)/10), k₂₀ ≈ 0.15 (white), 0.3 (amber), 0.5 (red) per day. The part of the consumed oxygen that SO₂ does not "catch" (when free SO₂ is low) oxidises ethanol to acetaldehyde (1 mg O₂ → 1.4 mg) and browns the phenolics. An illustrative model.
Copper sulphate (CuSO₄·5H₂O) contains 25.45% copper. Copper binds H₂S and mercaptans into insoluble sulphides; it does not act on disulphides. In the EU the limit for residual copper is 1 mg/L and the maximum copper sulphate dose is 1 g/hL. Copper sulphate is poisonous: work with gloves and accurate scales; excess copper makes wine hazy.
Oxygen plays a double role in the life of wine. At the start of fermentation yeast needs it to build sterols and cell membranes. In red wine, moderate oxygen softens tannins and stabilises colour: anthocyanins and tannins are linked by acetaldehyde "bridges" into stable polymeric pigments, which is the essence of barrel ageing and micro-oxygenation. But excess oxygen harms wine: white wine turns brown, fruit aromas fade, and smells of apple, hazelnut and "sherry" (acetaldehyde) appear; vinegar bacteria turn alcohol into vinegar.
The opposite of oxygen is a reductive environment: when wine gets no oxygen at all, especially on lees, sulphur compounds appear, with smells of rotten egg, cabbage and rubber. The winemaker's task is to find the balance between these two extremes.
Oxygen combines with wine compounds slowly on its own. Iron and copper ions speed up the reaction (as catalysts). The chain is as follows:
1. Oxygen oxidises phenolics (catechin, caffeic acid and others), forming quinones and hydrogen peroxide (H₂O₂).
2. In the presence of iron (the Fenton reaction), hydrogen peroxide oxidises ethanol to acetaldehyde.
3. Quinones combine with each other and with other compounds, forming brown pigments; they also capture thiols (the compounds that carry fruit aroma).
SO₂ breaks this chain: it reduces the peroxide and the quinones and is itself turned into sulphate. So as long as the wine has enough free SO₂, oxygen mainly uses up SO₂ (≈ 4 mg SO₂ per mg of oxygen), not the wine's colour and aroma. When the SO₂ runs out, the damage begins.
The rate at which dissolved oxygen is consumed depends on the wine type and temperature. Red wine, with plenty of phenolics, consumes oxygen in a few days, white wine in weeks. Every 10 °C roughly doubles the rate:
$$\frac{dO_2}{dt} = -k\,O_2, \qquad k(T) = k_{20}\cdot 2^{(T-20)/10}$$
At 20 °C at most ≈ 8 mg/L of oxygen dissolves in wine (when saturated with air). A single "hit" cannot give more than this, but repeated operations add up.
Wine takes up oxygen continuously (through vessel walls, the closure, the headspace) and during operations. Operations are often the main source:
| Operation | Oxygen, mg/L |
|---|---|
| Open racking with splashing | 3–6 |
| Racking by siphon, from the bottom | 0.5–1.5 |
| Pumping | 1–2 |
| Filtration | 0.5–2 |
| Topping up | 0.2–0.5 |
| Bottling (without inert gas) | 1–4 |
| Storage: glass, steel, full | ≈ 0.1 per month |
| Storage: oak barrel 225 L | ≈ 2–3 per month |
At bottling oxygen enters in two ways: it dissolves in the wine as it is filled, and air remains in the headspace in the bottle neck. In a 750 mL bottle, 10 mL of air contains ≈ 2.8 mg of oxygen, i.e. ≈ 3.7 mg/L. So a bottle should be filled up to the neck, and for valuable wine the bottle is first flushed with CO₂ or nitrogen.
The "oxygen capacity", how much oxygen a wine tolerates without significant damage, depends on its phenolics: white wine ≈ 10–20 mg/L, amber ≈ 30–50, red ≈ 40–80 mg/L (over its whole life). These are approximate figures, but they show that white wine needs very little oxygen.
Hydrogen sulphide (H₂S) is produced by yeast during fermentation. Yeast reduces sulphate and sulphite to sulphide, which should be built into sulphur-containing amino acids (cysteine, methionine). When the must lacks nitrogen, the yeast does not have the "skeleton" to build the amino acids, and the sulphide is released as H₂S. H₂S is also increased by sulphur residue sprayed in the vineyard, stressful fermentation (high temperature, high sugar) and some yeast strains. After fermentation H₂S is released from the lees as dead yeast breaks down.
The perception threshold for H₂S is very low, ≈ 1 microgram per litre. If it is not removed in time, it turns into other compounds in the wine:
| Compound | Smell | Does copper help? |
|---|---|---|
| Hydrogen sulphide (H₂S) | Rotten egg | Yes; air contact also helps |
| Methanethiol, ethanethiol (mercaptans) | Cooked cabbage, onion, rubber | Yes |
| Dimethyl and diethyl disulphide | Garlic, rotten bulb, marsh | No: reduction is needed first |
| Dimethyl sulphide | Cooked corn, radish | No |
Disulphides form from mercaptans in the presence of oxygen. So a reductive smell should be dealt with as early as possible, at the H₂S or mercaptan stage. In a closed bottle, isolated from oxygen, disulphides can be reduced back to mercaptans, which is why a reductive smell sometimes "returns" a few months after bottling.
The copper ion (Cu²⁺) binds H₂S and mercaptans into insoluble sulphides and mercaptides, which have no smell:
$$Cu^{2+} + H_2S \rightarrow CuS\downarrow + 2H^+$$
By molar ratio ≈ 1.9 mg of copper is needed to remove 1 mg of H₂S, but H₂S is present in wine in micrograms, so the copper dose is very small: 0.1–0.5 mg/L. Copper sulphate (CuSO₄·5H₂O) is used, which contains 25.45% copper. In the EU its maximum dose is 1 g/hL (≈ 2.5 mg/L copper), and residual copper in wine must not exceed 1 mg/L. Excess copper makes wine hazy (copper casse), speeds up oxidation and is undesirable for health.
The dose is always chosen with a bench trial: different amounts of copper sulphate solution are added to wine in several glasses, after 15–30 minutes all are smelled, and the smallest dose that removed the smell is chosen. If the smell disappears only with copper and ascorbic acid together, the wine contains disulphides: ascorbic acid reduces them to mercaptans, which copper can then bind. Such treatment is difficult and needs a laboratory's help.
Protection from oxidation: vessels always full, as few operations as possible and closed ones (by siphon, from the bottom), free SO₂ at target, wine kept cool. Protection from reduction: enough nitrogen in the must (Chapter II, "Preparing the must"), moderate fermentation temperature, stopping sulphur sprays in the vineyard 3–4 weeks before harvest, quick removal of the wine from the gross lees, and periodic stirring (bâtonnage) when ageing on fine lees. During long skin contact in a qvevri, smell the wine regularly: if a reductive smell appears, rack it in good time.
1. In the "Oxygen budget" mode choose the wine type, the storage vessel and the storage time, and enter how many times you plan each operation before bottling. You will see how much oxygen the wine will take up in total, how much SO₂ it will use, what share of the wine's "capacity" this is, and which source is the largest. Try replacing open racking with a siphon: the difference shows at once.
2. In the "Oxygen consumption" mode enter how much oxygen entered the wine in one operation, the temperature and free SO₂. The chart shows how the dissolved oxygen disappears, how much SO₂ it uses, and how much acetaldehyde forms if there is not enough SO₂.
3. In the "Reductive smells and copper" mode choose the smell, the copper test result and when the smell appeared: you get the likely compound and advice. Below, for a bench trial, enter the copper sulphate solution strength and the sample volume: the chart shows how many millilitres of solution to put in each glass. For the chosen dose you get the amount of copper sulphate for the whole wine and the residual copper.
1. Take two identical glasses and pour the same wine into both.
2. Put a clean, polished copper wire or piece of copper pipe (without lacquer or tarnish) into one glass. The other is the control.
3. After 5–10 minutes smell both. If the smell has gone or is much weaker in the copper glass, the cause is H₂S or mercaptans.
4. If it has not gone, add a few crystals of ascorbic acid and copper to a third glass: if it disappears now, the wine contains disulphides.
1. Make a 0.1% copper sulphate solution: 1 g of CuSO₄·5H₂O in 1 litre of distilled water (or 0.1 g in 100 mL).
2. Pour an accurately measured volume of wine (e.g. 500 mL) into 5 identical glasses and use a 1 mL syringe to add the calculated amount of solution. Add nothing to the first glass: it is the control.
3. Stir, and after 15–30 minutes smell and taste. Choose the smallest dose that removed the smell.
4. Add the same dose to the whole wine and stir well. After 1–2 weeks rack the wine off the sediment or filter it.
• Rack wine by siphon, with the end of the hose at the bottom of the receiving vessel, so the wine does not splash.
• There is one exception: if a young wine has a faint H₂S smell, open, splashing racking often removes it completely.
• Red wine needs moderate oxygen: an open first racking is beneficial. Protect white wine from oxygen as much as possible.
• At bottling fill the bottle to the neck; the headspace under the closure should be no more than 1–2 cm.
• Copper sulphate is poisonous: keep it in a labelled bottle, away from children. Its dose is measured in milligrams: do not add it by eye.
1. Which operation gives wine the most oxygen?
In open, splashing racking the wine becomes almost saturated with air: 3–6 mg/L of oxygen each time. By siphon from the bottom 0.5–1.5 mg/L; in full glass ≈ 0.1 mg/L per month.
2. Why does red wine withstand oxygen better than white?
Phenolics are an oxygen "buffer". White wine has few phenolics, so 10–20 mg/L of oxygen is already enough to brown its colour and strip its aroma.
3. A wine smells of cabbage. In the copper test the smell disappeared. What does this point to?
Copper binds H₂S and mercaptans into insoluble sulphides. If the smell does not disappear with copper, the cause is something else, or disulphides, which first need reducing with ascorbic acid.