Sulphur Dioxide

©Copyright Ben Rotter 2001-2011


1. Introduction
2. SO2 Production by Yeast
3. Sodium and Potassium Salts
4. Forms and Functions of Sulphur Dioxide in Wine
5. SO2 Binding
6. The Properties of SO2
7. Free SO2 and pH
8. SO2 and Temperature
9. Sensory Threshold
10. SO2 Loss
11. SO2 and Oxidation
12. Hyperoxidation
13. Accounting for SO2 Binding
14. Testing for SO2 (Ripper and AO methods)
15. Removing Free SO2
16. Adding SO2: Practical Considerations
17. Typical SO2 Additions
18. Storage and Purity
19. Stock Solutions
20. Campden Tablets
21. Sulphur Wicks and Rings
22. References

1. Introduction

Sulphur dioxide, often abbreviated to sulphite or SO2, has been used in winemaking since Roman times. It is used extensively in modern winemaking, predominantly for its suppression of yeast and bacterial action, and its anti-oxidant properties. It is possible to make wine successfully without using sulphites, but this generally results in reduced ageability, consistency and biological stability.

This article outlines the properties, forms, and uses of sulphur dioxide. Attention is given to the SO2 in general, the different forms of SO2 that exist in wine, and the issues of pH, temperature, SO2 binding, oxidative protection, SO2 removal, practical aspects of addition, hyperoxidation, SO2 testing, storage, stock solutions, sulphur wicks and Campden tablets.

Both theoretical and general practical aspects are presented in this article. Winemakers who do not wish to concern themselves with the more theoretical side to SO2 use may choose to skip such sections.

2. SO2 Production by Yeast

Sulphur dioxide is a natural by-product of yeast during fermentation [Zang and Franzen, 1966]. Usually less than 30 mg/l is formed. Zang and Franzen [1967] observed levels of 7-128 mg/l in twenty German wines, while Würdig and Schlotter [1967] reported 13-114 mg/l produced in twenty ferments. Heinzel et al. [1976] found levels ranging 6-296 mg/l, and levels ranging 3.2-640 mg/l under aerobic conditions. Levels over 100 mg/l have also been reported by Rankine and Pocock [1969], Eschenbroch [1974], Dott et al. [1976], and Suzzi et al. [1985]. The amount produced depends on various factors, including the yeast strain and the wine environment [Rankine, 1968; Eschenbruch, 1974; Romano and Suzzi, 1993; Würdig and Schlotter, 1967; Dittrich and Staudenmayer, 1968; Rankine and Pocock, 1969]. The production of SO2 by yeast tends to be higher in musts with a low level of suspended solids [Liu and Gallander, 1982]. Eschenbroch and Bonish [1976] found that pH had an influence on the SO2 production by some strains, but not by others.

The disadvantages of using SO2 tend to be limited to its excessive use. For example, high concentrations of SO2 have an offensive odour and taste, and the formation of H2S and mercaptans under extended yeast lees ageing can increase at higher SO2 concentrations. A small percentage of the human population is allergic to SO2, and vulnerable individuals may experience asthmatic attacks when exposed to very low levels (~1 mg) of SO2. This is, however, uncommon. A number of people believe they suffer from an SO2 allergy when this is in fact not the case. Individuals who to not exhibit an allergic reaction to sulphured packaged dried fruit are unlikely to be allergic to SO2.

3. Sodium and Potassium Salts

Two salt forms of sulphite are generally used in winemaking: potassium metabisulphite (K2S2O5) and sodium metabisulphite (Na2S2O5).

The molecular weight of sodium metabisulphite is 190.2 and that of potassium metabisulphite is 222.4, whereas that of sulphur dioxide (SO2) is 64.1. The salts dissociate giving two moles of SO2 for each mole of the salt. Thus, the SO2 content of sodium metabisulphite is 2 x 64.1/190.2 = 67.4% and that of potassium metabisulphite is 2 x 64.1/222.4 = 57.6%.

Table 1. SO2 content in metabisulphite salts
SaltSO2 content
Sodium metabisulphite67.4 %
Potassium metabisulphite57.6 %

Winemakers generally prefer to use the potassium form for sulphite additions, since this increases the level of potassium in the wine which may later help to precipitate tartrates during cold stabilisation. Some believe the sodium form can contribute a `salty' flavour to wine.

4. Forms and Functions of Sulphur Dioxide in Wine

4.1. Dissociation of Forms

Potassium or sodium metabisulphite dissociate in water to potassium ions (K+) and singly ionised bisulphite, (HSO3)-.

Metabisulphite dissociates in the following way to form these fractions:

K2S2O5 + H2O ===> 2K+ + 2(HSO3)-

Sulphur dioxide is a bifunctional acid, and dissociates into three fractions. The quantity of each of these fractions depends on the thermodynamic constants and the pH. The dissociation is almost instantaneous.

The three fractions are molecular SO2 (SO2), sulphite (SO32-), and bisulphite (HSO3-). Dissociation of the various fractions is almost immediate.

Since wine is acidic, hydrogen ions are present (H+) and the bisulphite (HSO3-) can then transform to sulphur dioxide:

HSO3-+ H+ <===> H2O + SO2
singly ionized bisulphite + hydrogen ionwater+ unionized (molecular) sulphur dioxide


HSO3- + H2O <===> H+ + SO32-
singly ionized bisulphite + waterhydrogen ion+ doubly ionized sulphite

Thus, the relationships of the forms of SO2 in wine are shown completely by:

H2O + SO2 <===> H+ + (HSO3)- <===> 2H+ + SO32-
water + molecular sulphur dioxide hydrogen ion + bisulphite hydrogen ion + sulphite

The amount of free SO2 that comprises each fraction (bisulphite, sulphite, and molecular) is determined by the pH. Figure 1 shows the distribution of the different species for various pH values.

Figure 1. The various species/forms/fractions of SO2 at various pH values.

4.2. Functions of the Different Forms

4.2.1. Bisulphite (HSO3-)

Bisulphite is the predominant form of free SO2 at wine and fruit juice pHs.
It causes the inactivation of polyphenol oxidase (PPO) enzymes and the binding and/or reduction of brown quinones in juice. (PPO enzymes are the enzymatic catalysts which cause oxidative browning of juice.) For more information, see section 11.1. Bisulphite is a successful extractive of anthocyanin (the predominant colouring matter in red fruits), yet it also bleaches colour and slows anthocyanin polymerisation reactions with other phenols.
Bisulphite possesses a low, and largely insignificant, antiseptic affect on yeasts (roughly twenty times less active than SO2 in wines with reducing sugars). It is odourless, but has a salty, bitter taste.

4.2.2. Molecular (or active) SO2

Molecular SO2 exists as either a gas or as single molecules in juice and wine. It is the most important form of SO2 in wine. It is responsible for antimicrobial activity [Rahn and Conn, 1944; Rhem, 1964; Macris and Markakis, 1974; Beech et al., 1979; King et al., 1981]. Rehm and Wittman [1962] found the antibacterial activity of molecular SO2 to be 500 times greater than bisulphite (HSO3-). It also possesses antioxidant activity (see Section 11, "SO2 and Oxidation"). It is volatile, and is responsible for the odour and sulphurous taste of SO2.

4.2.3. Sulphite (SO32-)

At typical wine pHs, the quantity the sulphite form of sulphur dioxide is minute and its reaction with oxygen is very slow. It is the only form which reacts with oxygen directly. It is odourless and tasteless at the concentrations typically used in juice and wine.

5. SO2 Binding

5.1. General

A portion of the SO2 added to wine will become bound with compounds in the wine. This portion is called "bound" (or "combined" or "fixed") SO2. The remainder is called "free" SO2 (FSO2). "Total" SO2 (TSO2) is the sum of free and bound SO2. Figure 2 represents this graphically.

It is the bisulphite form of SO2 which binds with other compounds. The bound SO2 compounds are therefore often termed "bisulphite addition products" and are sometimes referred to as "hydroxy-sulphonates". Unstable SO2-bound products may provide a reserve that feeds free SO2 when it subsides through oxidation or vaporisation. However, the degree of this re-partitioning is dependent on the binding kinetics of individual SO2-bound products and may not be practically significant in the majority of wines.

Figure 2. Free, bound and total forms of SO2. (Not proportionally to scale.)

SO2 that becomes bound is no longer available as free SO2. Since it is free SO2 that exhibits antimicrobial activity and oxidative protection, it is important to consider the amount of SO2 that will become bound.

Bound SO2 does not possess antiyeast activity, but the fractions of bound SO2 that are bound to acetaldehyde and pyruvic acid exhibit antibacterial action (this action is 5 to 10 times weaker than that of free SO2, though it is often present at 5 to 10 times the concentration of free SO2).

5.2. Compounds that bind

Binding compounds include carbonyl compounds, ketonic acids, sugars, quinones, anthocyanins, and others [Burroughs and Sparks, 1973a, 1973b, 1973c]. These are dealt with individually below.

5.2.1. Carbonyl binding

SO2 combines with the carbonyl groups of aldehydes [Hennig and Burkhardt, 1960a, 1960b]. Carbonyl compounds represent the predominant compounds that bind to SO2. By far the most significant carbonyl compound involved in SO2 binding is acetaldehyde.

Acetaldehyde (CH3CHO), also called ethanal, is a natural intermediate product during fermentation. It is oxidised ethanol, and is the compound that gives sherry its characteristic (oxidised) aroma. The main factors influencing acetaldehyde concentration are the yeast strain, the juice thiamine content, and the amount of SO2 added to the must. The increased presence of free SO2 in wine during fermentation also increases its production [Lafourcade, 1955]. Increased acetaldehyde concentrations are undesirable, since the aroma of acetaldehyde is generally not considered favourable in wine (at least at significant levels). Additionally, any addition of SO2 to a fermenting wine will immediately become bound with acetaldehyde, rendering it ineffective for its intended purpose. It is therefore best to avoid making SO2 additions during fermentation.

Bisulphite binds with the carbonyl oxygen atoms of acetaldehyde readily, making acetaldehyde the compound that binds most quantitatively with SO2. For example, Ough [1959] found that 100% of the BSO2 was bound to acetaldehyde in 12 out of the 17 wines they studied. Though Rankine [1966] examined wines in which not all bound SO2 could be accounted for by acetaldehyde binding. Kerp [1904a, 1904b, 1904c] noted that SO2 is bound almost entirely by acetaldehyde, except in heavily sulphited wines in which the excess SO2 was bound by sugars. Analysing data from 8 wines from Peynaud and Lafourcade [1952], shows that 100% of all acetaldehyde was bound to SO2. The analysis of data from 16 Californian wines [Ough, 1959], however, suggests that most of the wines in that study had 60-100% of their acetaldehyde bound to SO2.

The reaction for the binding of acetaldehyde and SO2 is:

CH3-CHO + HSO3- <==> CH3-CHOH-SO3-

and is independent of temperature (within normal ranges). Each milligram of acetaldehyde will bind with 1.45 milligrams of SO2. (The acetaldehyde-bisulphite compound is more correctly described as acetaldehyde-alpha-hydroxy sulphonate.)

Typical acetaldehyde concentrations in a newly fermented table wine are less than 75 mg/l. Typical concentrations in table wine range from 20 to 400 mg/l [Rankine, 1995]. The sensory threshold has been reported as 100-125 mg/l [Berg et al., 1955].

Following fermentation, SO2 might be added in sufficient concentrations for acetaldehyde to be completely bound [supported by Kielhöfer, 1963; Blouin, 1963 and 1966]. In such cases, a typical newly fermented table wine will bind a maximum of 110 mg/l SO2, though this figure may range from 30 to 580 mg/l. These figures will vary widely from wine to wine. Additionally, binding is quantitatively less, and the rate of binding slower, the lower the pH. Hennig [1943], however, believed that some acetaldehyde was beneficial for the development of wine bouquet.

Since SO2 will so readily bind with acetaldehyde, thus rendering it ineffective as an antimicrobial and antioxidative agent, it is suggested that acetaldehyde concentrations be kept to the minimum possible. This can be achieved by low sulphiting of the must, low pH, low fermentation temperature, and minimal exposure to air.

5.2.2. Ketonic acid binding

Ketonic acids (alpha-ketoglutaric acid, pyruvic acid, glutaric acid, keto-2-gluconic acid, diketo-2,5-gluconic acid, galacturonic acid) also bind to SO2. Increased ketonic acid levels may be the result of nutritional deficiency, especially a lack of vitamins due to mold infected fruit [Burroughs and Sparks, 1973b] or musts which have experienced ion exchange. Different yeast strains also produce different levels of alpha-ketonic acids [Farris et al., 1982 & 1983]

Pyruvic acid forms during fermentation. It has also been shown to decrease slowly once having formed, and the addition of thiamine has been shown to reduce its formation [Peynaud and Lafon-Lafourcade, 1966; Delfini et al., 1980].

The average concentrations of pyruvic acid and alpha-ketoglutaric acid in wine range 10-500 mg/l and 2-350 mg/l, respectively [Usseglio-Tomasset, 1989]. Typical ranges are 0-100 mg/l and 15-40 mg/l, respectively [Rankine, 1966].

5.2.3. Sugars binding

Sugars (arabinose, mannose, galactose, glucose, keto-5-fructose, xylsosone) bind to SO2. Polysaccharides also bind with SO2.

Binding to glucose is significant in juice and approximately 50% of added SO2 may become bound (at addition levels of 50-100 mg/l). Glucose has a low binding rate (0.8 mg SO2 per gram glucose in the presence of 100 mg/l free SO2). According to Braverman [1963], fructose does not form an addition product with the bisulphite form of SO2. In any case, SO2 binding with fructose is low, and is certainly less than that with glucose. Arabinose binds more readily, though its concentrations in wine are usually low. Saccharose does not bind to a significant level.

Data plotted in Figure 3 show the percentage of added SO2 which has become bound in concentrated orange, lemon and grapefruit fruit juices. (Data from Tressler et al. [1980].) As is expected, juices with a higher sugar content bind with more SO2 due to their increased sugar levels.

Figure 3. Percentage of added SO2 which becomes bound in concentrated fruit juices

5.2.4. Dicarbonyl group molecule binding

Dicarbonyl groups (glyoxal, methylglyoxal, hydroxypropanedial); gluconic and galacturonic acid, glyoxylic acid, oxaloacetic acid, glycolic aldehyde, glyceric aldehyde, dihydroxyacetone, acetoine, diacetyl, 5-(hydroxymethyl) furfural) bind with SO2.

5.2.5. Other binding

Bisulphite also binds to yeast, bacteria, and other protein and cellular particulates. This means SO2 additions are more effective in clarified juice. Binding can also occur with oxidation products of phenols and ascorbic acid. Phenolic compounds (proanthocyanic tannins, and particularly caffeic and p-coumaric acid) also bind reversibly with SO2 [Hennig and Burkhardt, 1960a, 1960b]. The binding is temperature and pH dependent. SO2 binds specifically with the four position carbons of monomeric anthocyanins of reds resulting in colour bleaching. With time, red wine pigments bind with non-SO2 compounds in wine and become less easily bound to SO2.

5.3. Binding kinetics

Binding is not instant. Whilst it is fastest within the first 24 hours of SO2 addition, it takes days before full binding is complete. Four to five days are typically required before binding ceases, during which time a slow decrease in free SO2 occurs, accompanied by a corresponding increase in bound SO2. After this time, equilibrium is realised and any further decreases are due to oxidation or vaporisation.

The rate of binding is dependent on the dissociation constant, K, for the individual binding reaction. The lower the value for K, the more favoured the formation of a bisulphite addition product [see Burroughs and Whiting, 1960]. As an example, acetaldehyde has a low K value compared to many other compounds in wine. The binding of SO2 to acetaldehyde is strong and rapid. For example, at a pH of 3.3, 98% of acetaldehyde was found to bind within 90 minutes of addition, and total combination was complete within 5 hours. In one study [Joslyn and Braverman, 1954], 90-95% was found to bind after just 2 minutes.

A number of K values are given in Table 1.

Table 2. Dissociation constants for SO2 binding with select compounds
SubstanceSource 1Source 2Source 3Source 4Source 5Source 6
Formaldehyde 1.2 × 10-7
Acetaldehyde 2.5 × 10-65 × 10-4 1.5 × 10-61.5 × 10-61.5 × 10-6
alpha-ketoglutaric acid 8.8 × 10-48.8 × 10-45 × 10-4
Benzaldehyde1 × 10-4
Acetone 3.5-4.0 × 10-33.8 × 10-3
Furfural 7.2 × 10-4
Chloral 3.5 × 10-2
Arabinose 3.5 × 10-2
Glucose 2.2 × 10-1
Pyruvic acid 0.3 × 10-31.4 × 10-44.0 × 10-44.0 × 10-43 × 10-4
Glucose 9 × 10-1 6.4 × 10-16.4 × 10-16.4 × 10-1
Fructose 1.5
Sucrose 5.4

Source 1 - Kolthoff and Stenger [1942] at 25 C
Source 2 - Blouin [1966] at 20 C
Source 3 - Burroughs and Sparks [1973a] and Beech et al. [1979] at pH=3
Source 4 - Burroughs and Whiting [1960] at pH 3-4, 20 C, 50 mg/l FSO2, in cider
Source 5 - Rankine [1966]
Source 6 - Usseglio-Tomasset [1989]

The value of the dissociation constant, K, increases with increasing temperature. For example, the K value for acetaldehyde increases by 5 times from 25°C (77°F) to 37.5°C (99.5°F) [Kerp, 1903; Kerp and Bauer, 1904 and 1907a,b].

It is difficult to generalise binding kinetics for all wines. Nevertheless, Blouin [1966] found that, at a free SO2 of 20 mg/l, substances with a K value less than 3 × 10-6 bound completely with SO2, whilst substances with a K value equal to or larger than 3.1 × 10-2 bound less than 1%. The presence of metals was also found to increase binding.

5.4. Relationship between total and free SO2

Molecular SO2 provides the predominant protective qualities of SO2 in wine. Since the molecular SO2 present is only a portion of the free SO2, it is important to take into consideration the fact that a fraction of SO2 additions will become bound and will no longer remain as free.

The relationship between the amount of added SO2 and the amount of SO2 remaining free is complex. It is clear, however, that it is governed by the total SO2 content of the wine and the ability of the individual wines' compounds to bind with SO2. The higher the total SO2 content, the less that a further SO2 addition will bind. The relationship between free SO2 and bound SO2 is shown in Figure 4. The rate of binding can be seen to decrease as the free SO2 concentration increases.

Figure 4. Relationship between free and bound SO2.

The exact relationship between free and bound SO2 will vary from wine to wine, but can be determined for an individual wine by making a range of SO2 additions to sample of the wine (for e.g., sample 1 has 10 mg/l SO2 added, sample 2 has 20 mg/l SO2 added, etc). The free SO2 is then measured in the samples some time later (4-5 days later is safest). An understanding of the wine's SO2 binding response at that particular point in it's life can then be determined. Such a task is, however, laborious and time consuming. Instead, many winemakers assume an empirical law of binding. This simply estimates the portion of added SO2 which will become bound on addition.

5.5. Approximate rules for SO2 binding

Many winemakers assume that about 50% of their SO2 addition to a wine becomes bound when the total SO2 is below 30-60 mg/l. After this level, added SO2 is generally considered not to bind, providing free SO2 almost exclusively. (Though some winemakers assume that thereafter some SO2 does become bound (usually about 30%).)

Peynaud notes that roughly one third (33%) of added SO2 becomes bound [Peynaud, 1984, p.271,250] under a free SO2 content of 100 mg/l.

Margalit [1996, p.268] cites the SO2 binding data of Schaeffer [1987], who found that 70-85% of SO2 became bound when added to newly fermented Gewurztraminer wine. Margalit [1990, p.26] further notes that for healthy fruit, around half of any SO2 addition will become bound when the total SO2 concentration is under 50-60 mg/l. Above 50-60 mg/l total SO2, any addition is considered to contribute to free SO2 in its entirety, i.e. no binding occurs.

Low fermentation temperatures, anaerobic fermentations, addition of ammonium salts, and use of non-ketogenic yeast strains may all help to minimise potential addition products and hence minimise SO2 binding [Peynaud and Lafon-Lafourcade, 1966].

6. The Properties of SO2

6.1. Antioxidant

SO2 protects both must and wine from excessive oxidation.

6.2. Antienzymatic

Sulphur dioxide inhibits oxidation enzymes (enzymatic catalysts of oxidation such as tyrosinase and laccase) and destroys them with time. It inhibits the polyphenol oxidase enzyme responsible for catalysing oxidative reactions in juice. The oxidative protection of a must is sustained by this mechanism before fermentation begins. Its inclusion in must will therefore increase the amount of oxygen available to yeast in their growth phase. The use of SO2 can help to avoid oxidasic casse from rotten fruit.

6.3. Taste

Acetaldehyde is the compound which gives sherry its characteristic oxidised (or maderised) aroma. Similarly, a small amount of wine left out in a glass overnight will show an aroma dominated by acetaldehyde. A common fault due to excessive oxidation is the presence of high concentrations of acetaldehyde. SO2 will bind with acetaldehyde, essentially removing its volatile presence and resulting in a wine with a "fresher" aroma.

The addition of SO2 at crush will increase the extraction of flavonoid phenols [Singleton et al., 1980]. These compounds will contribute to bitterness and astringency. The absence of SO2 in must will increase the oxidative polymerisation [Ough and Crowell, 1987] and precipitation of phenols.

The presence of SO2 to the must may additionally cause increased extraction of phenolics during maceration.

Wines low in SO2 are believed to have softer palates, whilst high levels increase the harshness of wine. Excessive levels cause wines to have pungent, sulphurous aromas.

6.4. Fermentation

At low levels of 5-10 mg/l, SO2 delays the onset of fermentation, but later speeds up the multiplication of yeasts and their transformation of sugars [Peynaud, 1984]. Higher additions, however, result in increased fermentation delays [Yang, 1975]. This is attributed to the SO2's destruction of fungicides, which are toxic to yeast. The delay in the initiation of fermentation assists in pre-fermentation juice settling of musts. It also allows for a greater level of dissolved oxygen in the must, due to the prevention of such oxygen being used in enzymatic oxidation reactions. These higher dissolved oxygen levels provide a healthier environment for the yeast.

In musts neither inoculated with cultured yeast nor with SO2 added, wild yeast species such as the genii Kloeckera and Candida are usually present during the early stages of fermentation, but are soon dominated by the genus Saccharomyces which take a stronger hold over the fermentation. SO2 can inhibit the former strains, allowing for the latter strains to dominate fermentation from the on-set. In modern winemaking this is generally deemed more preferable. However, there are winemakers who believe that the limited influence of wild strains can enhance a wine's character.

6.5. Colour

Use of SO2 during crush/fermentation causes an increase in colour [Harvalia, 1965] through the extraction/solvency of anthocyanins and polyphenols from fruit tissues, though at normal doses the colour increase is aesthetically insignificant [Amerine and Joslyn, 1951]. (This is because the anthocyanidin-SO2 compound is more soluble in water-ethanol than anthocyanidin alone.) Wines fermented with SO2 have also been found to better retain colour [Berg and Akiyoshi, 1962]. Excessive amounts of SO2, however, cause colour bleaching, though this discolouration of red pigments is reversible.

Anthocyanin pigments bind readily with bisulphite (HSO3-). In this reaction, the coloured anthocyanin cation binds with the bisulphite anion to form colourless anthcyanin-4-bisulphite [Jurd, 1964]. This reaction has been shown to be 85% complete with the addition of 15 mg/l SO2 [Timberlake and Bridle, 1976]. However, polymeric anthocyanins are resistant to SO2 and contribute to a significant proportion of red wine colour [Burroughs, 1975].

6.6. Browning

Sulphur dioxide reduces browning by obstructing polyphenol oxidase (PPO) enzymes. These are the enzymatic catalysts which cause oxidative browning of juice. When an apple is freshly cut and the flesh begins to turn brown, for example, this is due to the activity of PPO enzymes. The reaction is as follows:

Polyphenol oxidase (PPO) activity has been shown to be reduced by more than 90% by the presence of 50 mg/l SO2 [Dubernet and Ribéreau-Gayon, 1973; Amano et al., 1979].

It seems that the bisulphite (HSO3-) form is responsible for this, and that this occurs by irreversible structural modification rather than binding inhibition [Sayavedra-Soto and Montgomery, 1986].

6.7. Antimicrobial

At low concentrations, SO2 inhibits the development of microorganisms. At high concentrations, it can destroy a proportion of the microbial population. Molecular SO2 is the form responsible for antimicrobial action [Rahn and Conn, 1944; Rhem, 1964; Macris and Markakis, 1974; Beech et al., 1979; King et al., 1981].

Bound SO2 also possesses antimicrobial activity, though this is limited. The antimicrobial activity of bound SO2 depends on the compound that the SO2 is bound to [Rehm and Wittman, 1962]. For example, acetaldehyde, pyruvate, and acetone have a significant inhibition effect, whilst glucose has only a slight inhibition effect. Generally, the antimicrobial activity of bound SO2 is not significant.

Different yeast and bacteria strains have different levels of tolerance to SO2 [Cruess, 1912]. A number of studies have attempted to determine the levels required for inhibition or death for numerous strains. One study found a 1000-fold reduction in the number of viable cells of a Brettanomyces species, certain LAB species, and other wine spoilage organisms within 24 hours at a molecular SO2 concentration of 0.8 mg/l [Beech et al., 1979]. Another found total microbial inhibition in musts at 4 mg/l molecular SO2 [Delfini, 1984]. A classic study by Beech et al. [1979] assessed the SO2 required to reduce non-growing yeast and bacterial populations by 10,000 viable cells/ml over a 24 hour period in 10% ethanol buffered solutions. They found that 0.825 mg/l molecular SO2 was required for one Saccharomyces cerevisiae strain, 0.825 mg/l for a Brettanomyces strain, 1.50 mg/l for a Zygosaccharomayces bailii strain, and 4 mg/l for a Lactobacillus plantarum strain. Based on these figures it would seem that a level of 0.8 mg/l molecular SO2 is sufficient for the suppression of the majority of yeast and bacteria strains.

It should be noted, however, that strains can build up resistance to SO2. Older cultures tend to have more resistance to SO2 [Schimz, 1980; Katchmer, 1990].

Bacteria are more susceptible to SO2 than yeasts and are considered separately below.

6.8. Antiyeast

Molecular SO2, and to a lesser extent bisulphite (HSO3-), inhibit yeast. Free SO2 essentially has an antiseptic effect [Kielhofer, 1963], and the growth of Saccharomyces has been shown to be related to its concentration [Ingram, 1948].

6.8.1. Resistance adaptation

As stated above, different yeast strains are resistant to SO2 [Porchet, 1931] to varying degrees. Some may tolerate 700 mg/l free SO2 or more.

Yeast may also adapt to an SO2 environment and become resistant to SO2. Certain yeasts have been shown to permanently adapt to 10-12 times the SO2 concentration that the parent strain could tolerate [Scardovi, 1951, 1952, 1953]. Delfini [1988, 1989, 1992a] demonstrated that a variety of yeast strains (S.cerevisiae, S. ludwigii, Zygosaccharomyces baillii, and Schizosaccharomyces japonicus) could develop successive permanent (inherited) resistance to SO2 to a final level of 9.2-11.5 mg/l molecular SO2.

It is therefore important to limit SO2 additions, and to avoid adding successive doses, as this may result in increased SO2 resistance by the yeast strain.

6.8.2. Growth

Yeast growth exhibits an extended lag phase in the presence of SO2, but this is usually followed by normal growth following the end of the lag phase [Schanderl, 1959].

SO2 is more effective on yeasts in their resting/sporulating phase, since binding with aldehydes may occur latter.

6.8.3. Complete and partial inhibition

Scardovi [1951] showed that total S. cerevisiae strain cell death occurred with 4 mg/l molecular SO2 for non-resistant variants, whilst 40 mg/l was required for resistant strains. S. cerevisiae has been shown to be sensitive to 0.5-0.9 mg/l molecular SO2, with complete inhibition occurring at >0.5-1 mg/l [Beech, 1979]. In another study, approximately 30% cell death occurred at 18 mg/l molecular SO2, whilst 70% death occurred at 42 mg/l [Farkas, 1988]. SO2 can also reduce the viability of a yeast inoculum. One study found that 15-20 mg/l free SO2 reduced the population from 106 to 104 cells/ml [Lehmann, 1987]. Nevertheless, fermentation has been seen to commonly occur (evidenced by a 0.5-1% by volume alcohol production) in musts with as much as 2000 mg/l free SO2 [Delfini, 1984].

Marcis and Markakis [1974] showed that 1.3 mg/l molecular SO2 was required to eliminate viable yeast cells in a medium. Another study showed that 1.56 mg/l was required [King et al., 1981]. Minarik [1978] found that 6.4 mg/l was required in juice, while Beech et al. [1979] found that 0.825 mg/l was required in a model wine solution, and Sudraud and Chauvet [1985] suggested 1.5 mg/l be used following fermentation and 1.2 mg/l be used during storage to prevent refermentation of residual sugar.

It may be generally accepted that 4-5 mg/l molecular SO2 can cause total inhibition of S. cerevisiae.

6.8.4. Time dependence

Total yeast death is also time dependent. Yeast uptake of SO2 is rapid, and can be complete within 3 minutes [Macris and Markakis, 1974]. It may, however, require longer periods of time for SO2 to become lethal. One study [Delfini, 1981] found total inhibition occurred at 0.29 mg/l molecular SO2 for Kloeckera apiculata, 0.67 for Pichia vini, and 1.59 for Candida vini. These concentrations became lethal after 24 hours of exposure with a cell population of 106 cells/ml. Macris and Markakis [1974] found that a population reduction of 90% took 83 minutes with 0.025 mg/l molecular SO2. In spite of these findings, Uzuka and Nomura [1986] found that, at 0.80 mg/l molecular SO2, over 50% of yeast viability was lost within 30 minutes. A similar reduction corresponds to 6 hours at 0.825 mg/l in the Beech et al. [1979] study and 20 hours in the King et al. [1981] study.

6.8.5. Yeast selective

To a certain degree, SO2 may be used as a yeast selector. At certain doses it promotes yeast selection by hindering the multiplication of non/low-alcohol producing yeasts such as apiculates, Torulopsis, and Candida more than that of elliptic yeasts [Romano and Suzzi, 1992]. Nevertheless, Heard and Fleet [1988] showed that apiculated yeasts (Kloeckera and Hanseniaspora) grew to substantial populations (106-107 cells/ml) in a few days before receding.

6.9. Antibacterial

Lactic bacteria are sensitive to free and, to a lesser extent, bound SO2 [Fornachon, 1963].

The primary antimicrobial effect of SO2 is attributable to molecular SO2, at least up to pH 5 [Scardovi, 1951, 1952; Macris and Markakis, 1974]. Though there is evidence that bound SO2 can contribute to bacterial control [Bioletti, 1912; Rhem, Wallnofer and Wittman, 1965; Lafon-Lafourcade and Peynaud, 1974; Hood, 1983] and inhibit LAB growth [Fornachon, 1963]. This is because LAB consume acetaldehyde, which subsequently releases SO2 from the bound acetaldehyde-SO2 form [Osborne et al., 2000]. Mayer et al. [1975] found Leuconostoc oenos sensitive to levels of acetaldehyde-bound SO2 levels of 20-60 mg/l. Hood [1983] showed that just 6 mg/l of acetaldehyde-bound SO2 could inhibit the growth of Leuconostoc oenos, Leuconostoc brevis, and Pediococcus pentosaceus at pH 3.4.

In one study [Delfini and Morsiani, 1992], ten strains of Leuconostoc and four of Lactobacillus were found to cease growth above 0.5 mg/l molecular SO2. A Leuconostoc population of 2 × 106 cells/ml in a buffered synthetic medium died within 22 hours after addition of 0.84 mg/l molecular SO2. However, a number of Leuconostoc and Lactobacillus strains survived SO2 additions and resumed multiplication after a 10-60 day lag phase at molecular SO2 levels under 0.8 mg/l.

Acetic acid bacteria are also sensitive to SO2. Research conducted on Acetobacter aceti, A. liquefaciens, A. hansenii, A. pasteurianus and Gluconobacter oxydans showed that some strains were more sensitive than others. Molecular SO2 levels of 0.1-0.65 mg/l were required to effectively kill strains in juice over a 4 day period, depending on the individual strain [du Toit, 2000]. The production of VA was also found to inhibit the growth of yeast (Vin 13). Some acetic acid bacteria strains were found to produce SO2 binding compounds such as gluconic, 2 ketogluconic and 2,5 diketogluconic acids. The addition of SO2 before fermentation may therefore be of increased importance, since this will inhibit acetic acid development which in turn will prevent inhibition of yeast growth.

6.10. Overview

The complexities of SO2 inhibition/death on yeast and bacteria have not been exhaustively studied. The quantity of molecular sulphur dioxide required to inhibit specific micro-organisms depends on their individual environment and history. For example, pH impacts on yeast and bacterial growth irrespective of SO2 concentration, and should therefore be considered as a separate yet related influence. In the absence of such information, values utilised for protection must be obtained from the available data.

SO2 should not be used to stop fermentation directly, since the SO2 added will immediately become bound leaving it ineffective. If SO2 is used as a yeast inhibitor, its use should be in parallel with other techniques (such as low temperature, clarification, or sorbate) and even then should only be used when a sufficient reduction in the yeast population is attained.

SO2 remains an invaluable tool for inhibiting bacteria, which might otherwise spoil wine. It appears that levels of inhibition range from 0.5 to 0.825 mg/l molecular SO2, depending on the bacteria and strain. Based on this information, maintenance of 0.825 mg/l molecular SO2 would be a safe approach to controlling bacteria. However, levels used by winemakers to control biological stability are generally achieved through levels ranging 0.5-1.5 mg/l molecular SO2.

Currently, control of biological stability is generally achieved through levels ranging 0.5-1.5 mg/l of molecular SO2. A general opinion exists that 0.8 mg/l molecular SO2 provides sufficient protection for dry wines. (Some feel that a concentration of 0.6 mg/l is suitable for red must or wine, while 0.8 mg/l is suitable for white must or wine.) Wines with residual sugar might better be protected with levels ranging 1.5-2 mg/l. However, it should be kept in mind that these levels are rule-of-thumb and different yeast and bacteria under different conditions will act differently. Winemakers should be aware of this and arrive at usage levels suitable to their individual situation.

7. Free SO2 and pH

As mentioned previously, molecular SO2 is the principal form of sulphur dioxide responsible for anti-microbial activity. The amount of molecular free SO2 available is a function of pH. Thus, SO2 additions should be calculated with reference to pH.

Because the significant SO2 form responsible for antimicrobial action is molecular SO2, microbial growth control without measurement of free SO2 and pH is less meaningful and certainly less assured.

The relationship between pH and molecular free SO2 can be shown as follows:

[SO2] + [H2O] <===> [HSO3-] [H+]
K = [HSO3-] [H+] / [SO2]
[HSO3-] / [SO2] = K / [H+]
[H+] = 10-pH
K = 10-pKa
[HSO3-] / [SO2] = 10-pKa / 10-pH
[HSO3-] / [SO2] = 10pH - pKa

Since the sulphite form (SO32-) is almost insignificant at wine and must pH, free SO2 consists of the molecular form (SO2) and the bisulphite form (HSO3-),

Free SO2 = [HSO3-] + [SO2]
( Free SO2 - [SO2] ) / [SO2] = 10pH-pKa
( Free SO2 / [SO2] ) - 1 = 10pH - pKa
Free SO2 / Molecular SO2 = 10pH - pKa + 1
Free SO2 = Molecular SO2 * ( 10pH - pKa + 1 )

King et al. [1981] found the pKa's of SO2 in water to be 1.77 and 7.20. Tartar and Garretson [1941] and Schroeter [1966] reported values of 1.76 and 7.20. These values seem to have become the accepted values in water. The ionisation constants are affected by the ethanol concentration, the presence of sugars and other organic compounds and salts, and the temperature. Figure 5 shows various pK1 values for given ethanol concentrations and temperatures [data from Usseglio-Tomasset, 1989].

Figure 5. SO2 pK1 values for various ethanol concentrations and temperatures.

Usseglio-Tomasset and Bosia [1984] reported a pK1 value of 1.78 (at 19°C) (pK2 was 7.08 at 20°C) and noted that a pK1 value closer to 2.0 was more realistic in wine-like conditions. Thus, using a value of 1.81 might be more sensible.

However, the difference between using a pK1 value of 1.77 or 1.81 is minimal in terms of free SO2 values required (for any given molecular level). Thus, for most practical purposes, the importance in using an acutely accurate pK value is not high under normal winemaking conditions.

In the calculations below, a pK value of 1.81 is adopted. This is a widely used value. Thus, the equations relating molecular and free SO2 can be written as:

Molecular SO2 = Free SO2 / ( 10(pH - 1.81) + 1 )
Free SO2 = Molecular SO2 * ( 10(pH - 1.81) + 1 )

Alternatively, values obtained using the above equations are shown in Table 2 below.

Table 3. Free SO2 required for given molecular SO2 level
pH Free SO2 (mg/l) for given molecular SO2 level
0.6 mg/l 0.8 mg/l 2 mg/l
2.8 6 9 22
2.9 8 11 27
3.0 10 13 33
3.1 12 16 41
3.2 15 20 51
3.3 19 26 64
3.4 24 32 80
3.5 30 40 100
3.6 38 50 125
3.7 47 63 157
3.8 59 79 197
3.9 74 99 248
4.0 94 125 312

Figure 6. Free SO2 required for 0.6 and 0.8 mg/l molecular SO2.

Figure 7. Free SO2 required for 2 mg/l molecular SO2.

8. SO2 and Temperature

As temperature increases, free SO2 increases and bound SO2 decreases. (SO2 bound to acetaldehyde remains constant.) This is because increased temperatures cause partial dissociation of the bound SO2 form, resulting in increased free SO2 and hence increased molecular SO2 concentrations.

For example, a wine containing 68 mg of free SO2 at 0°C (30°F) will contain 85 mg at 15°C (57°F) and 100 mg at 30°C (84°F) [Peynaud, 1984]. Sudraud [1977] showed 64 mg/l of free SO2 at 16°C increased to 120 at 48°C and to 200 at 80°C, as BSO2 was released.

Sometimes, wines with high molecular SO2 levels are served cold to hide the sulphurous aroma they would exhibit with their high SO2 content.

9. Sensory Threshold

It is the molecular SO2 form which is responsible for the sensory threshold. Hence, the sensory threshold of SO2 depends on the pH and temperature. There exists considerable variation in threshold within the population. Nevertheless, the sensory threshold is generally considered to be around 2 mg/l molecular SO2.

10. SO2 Loss

SO2 can be lost from wine under a number of circumstances. Molecular SO2 is volatile and some is lost from both juice and wine through vaporisation to the air, especially if the juice/wine is agitated. This loss is higher in wines stored in barrels. However, the quantity lost in this way is usually negligible.
SO2 will be lost during alcoholic fermentation. This is partially through vaporisation with escaping carbon dioxide from the fermentation. At the end of fermentation it is common for a wine to possess zero to just a few milligrams per litre of total SO2, however significant deviations from this norm can be found.
Losses additionally occur through the oxidative protection of SO2. This is largely due to SO2 reacting with hydrogen peroxide to form sulphuric acid. Interactions of SO2 with quinones to form monosulphonates may also result in SO2 loss [Lu Valle, 1952].
SO2 is also lost in bottled wine [Ough, 1985]. Müller-Späth [1982] found that the total SO2 had dropped by 20-30% after 5 years at 12°C in two bottled white wines. The rate of total SO2 loss appears to be 2-3 times faster in reds than in whites [Ough, 1985]. Peynaud notes that SO2 losses in bottle are a few mg/l per year [Peynaud, 1984, p.271]. The causes for SO2 loss in bottle are numerous. SO2 vapour may be lost through the cork, but this is not substantial under normal temperatures of storage. Oxidation of the SO2 with oxygen in the bottle will certainly occur, but this reaction is very slow. Oxidation of the SO2 by formerly oxidised phenols will lead to the production of sulphate and a loss in total SO2. SO2 loss may also occur due to rearrangements in the reaction processes within the wine after long time periods, favouring redox/equilibrium reactions rather than kinetic reaction rates.

Given this information, it should be kept in mind that the total SO2 is not the same as the amount of SO2 that has been added to the must/wine, since some of the added SO2 will be oxidised irreversibly to sulphate and some lost through volatisation.

Whenever losses occur, the equilibrium between free and bound SO2 will re-establish, resulting in a small decrease in bound SO2.

11. SO2 and Oxidation

11.1. In must

Without the presence of SO2 in musts, juice undergoes enzymatic oxidation. The enzymatic oxidation of phenolic compounds governs oxidation over and above chemical oxidation reactions because of their much faster reaction speed. The oxidase responsible is polyphenol oxidase (PPO), and in the case of Botrytis infected fruit, laccase [Dubernet and Ribéreau-Gayon, 1973 and 1974].

PPO is also known as tyrosinase, catecholoxidase, catecholase, phenolase, phenoloxidase, and o-diphenoloxidase. Its activity depreciates with time and is usually completely inactive following fermentation. However, it is primarily responsible for oxidation in juice. SO2 is usually added to musts to inhibit the activity of (or destroy) oxidase enzymes and subsequently prevent oxidation. (For exceptions to this practise, see Section 12, "Hyperoxidation".)

The rate of oxygen uptake in juice is determined by the temperature, enzyme activity, phenolic concentrations, the substances "consumed" by the enzyme, and the competition between different substances for binding [White and Ough, 1973]. PPO activities vary widely even within fruits of the same variety [Traverso-Rueda and Singleton, 1973; Hooper et al., 1985].

However, in the absence of SO2, oxygen uptake is generally rapid. When first coming into contact with air, an SO2-free juice can exceed uptake of 2 mg oxygen/l/min. Uptake of dissolved oxygen (from a saturated juice state) can be complete in a white grape must within 4-20 minutes [Dubernet and Ribéreau-Gayon, 1974]. In an apple juice, also saturated with dissolved oxygen, consumption was complete within 1 hour (at pH 3.45) and most uptake took place within the first 10 minutes. The process was slower in orange juice at pH 3.57 with an initial rapid uptake of around 2 mg/l oxygen within the first hour, followed by around 0.4 mg/l uptake per hour thereafter. Uptake in lemon juice at pH 2.35 progressed at around 0.7 mg/l oxygen per hour. [Lüthi, 1953, 1954, 1960; Biedermann, 1956].

Upon the addition of SO2 to a must, oxygen consumption will tail off over a period of time (typically 1-6 minutes for SO2 additions of 10-100 mg/l) until it completely stops and the oxygen concentration remains constant [Dubernet and Ribéreau-Gayon, 1974]. This behaviour is shown in Figure 8.

Figure 8. Delay in halting oxygen consumption of must after SO2 addition.

Based on the fact that the effectiveness of the SO2 is delayed, and that such enzymatic oxidation is rapid, it is important to ensure that SO2 is added as soon as possible to prevent oxidation.

The addition of 25-75 mg/l SO2 to clarified juices has been shown to inhibit PPO activity by 75 and 97%, respectively [Dubernet and Ribéreau-Gayon, 1973; Amano et al., 1979]. A rough representation of these results is presented in Figure 9.

Figure 9. Reduced PPO activity with increased SO2.

Due to the binding of SO2 to fruit particulates, the required doses may in practice be 75-100 mg/l for unclarified must, and 30-50 mg/l for clarified juices.

Laccase (also known as p-phenoloxidase) is another important oxidative enzyme. It is found mostly in fruit infected with Botrytis cinerea and it causes rapid oxidation [Peynaud, 1984]. Less data exists on the inactivation of laccase by SO2. However, it appears more difficult to inactivate than PPO enzymes, and it consumes oxygen over longer periods of time than PPO enzymes. One study found that even with a free SO2 level of 150 mg/l, only a 20% reduction in activity occurred [Dubernet and Ribéreau-Gayon, 1973]. In general, laccase activity exists in fermented wines made from Botrytis infected fruit. Ascorbic acid is often used as an antioxidant in such situations.

11.2. In wine

The excessive oxidation of wines causes increased browning, increased production of stale-smelling aldehydes (for example, the oxidation of ethanol to acetaldehyde [Kielhofer and Würdig, 1960]), and ultimately leads to a loss of fruit/varietal character. Limiting oxidation (to a certain extent) is normal practice in most wines and exceptions to this are limited (e.g. Sherry, Madeira).

11.2.1. Mechanisms

In wine, oxidative enzymes no longer exist and the primary oxidative impact is through chemical oxidation.

The precise mechanisms by which SO2 protects wine from oxidation are not fully understood, nor widely agreed upon. Nevertheless, SO2 is deemed to possess antioxidative activity because it is preferentially oxidised over other compounds which, when oxidised, would lead to undesirable aromatic/flavour changes.

The direct reaction of SO2 with oxygen appears to be insignificant in wine. Oxygen reacts with polyphenols before it can be removed by SO2. Instead, the main antioxidant role of SO2 appears to be its reaction with hydrogen peroxide (H2O2) [Danilewicz, 2003]. The most reactive polyphenol grouping in grape (and apple) wine are the catechols (in this explanation catechols are used as a simple example to show the reaction of the most reactive wine polyphenol grouping). The following equation represents their reaction with O2 in wine:

1,2-benzenediol (catechol), through Fe(III) catalysis, reacts with O2 to form 1,2-benzoquinone and hydrogen peroxide [Wildenradt and Singleton, 1974].

Essentially, transition metals oxidise catechol (1,2-benzenediol) and these reduced metals are then reoxidised by oxygen. Indeed, work by Poulton [1970] suggests that an intermediate is formed in the presence of oxygen which then reacts rapidly with SO2 since, under model wine conditions, a half time of approximately 30 days was required for SO2 to consume the oxygen in a saturated solution. In a white wine, the half time was 1.2 days.

Hydrogen peroxide is a strong oxidising agent and will oxidise other wine compounds, potentially leading to an undesirable aroma. For example, ethanol may be oxidised to form acetaldehyde, which possesses a stale aroma:

H2O2 + ethanol ===> acetaldehyde + 2H2O

However, when SO2 is present, the bisulphite ion (HSO3-) is deemed to undergo nucleophilic attack by hydrogen peroxide (H2O2) with displacement of water and the formation of a peroxymonosulphite ion:

HSO3- + H2O2 ===> HOOSO2- + H2O

Acid catalyzed rearrangement subsequently results in the formation of a sulphate dianion:

HOOSO2- + H+ ===> SO42- + 2H+

Thus, SO2 removes the strong oxidising agent H2O2, preventing it from oxidising other wine compounds which would potentially lead to undesirable oxidised aromas. (This action is, however, competitive and formation of acetaldehyde has been shown to occur even in model wine solutions with high (177 mg/l) concentrations of SO2 [Wildenradt and Singleton, 1974].)

From the equations above it can be seen that, for every mole of oxygen uptaken, one mole of SO2 and one mole of a quinone will form.

Quinones are highly reactive with bisulphite and the reaction of 1,2-benzoquinone with bisulphite proceeds as follows:

However, some of the quinone will not react with bisulphite due to the fact that they condense with other compounds. If all the quinone reacted with bisulphite then a further mole of SO2 would be taken up, resulting in the uptake of two moles of SO2 for every mole of oxygen. However, since not all quinones will react in this way, the overall SO2 loss for every mole of oxygen should be between one and two moles of SO2. The antioxidant activity of bisulphite is primarily restricted to its reaction with hydrogen peroxide. The direct reaction of SO2 with oxygen seems to be a chain process, prevented by the reaction of scavanging phenols with SO2 [Danilewicz, 2003].

Wines made from botrytised fruit present an exception to the oxidation of wine being solely chemically driven. In this case, the enzyme laccase may be involved in the oxidation (see Section 11.1 for more).
Since SO2 does not prevent oxidation directly - it simply prevents undesirable oxidation reactions from taking place - measures to protect wine from oxidation should still be practised even when SO2 is used.

11.2.2. Reaction

Under ideal conditions, 2 moles of SO2 would be required to remove 1 mole of oxygen (molecular weight of 32 vs 64) in a direct reaction. However, as noted in the above section, the overall SO2 loss for each mole of oxygen should be between one and two moles of SO2, depending on whether quinones react with bisulphite (see Section 11.2.1). This seems unlikely since condensation with numerous compounds (including polyphenols) has been shown to occur. Thus, in practice, it appears more likely that 1 mole of SO2 will be lost for every 1 mole of oxygen consumed. It should, however, be stressed that the direct reaction of oxygen with SO2 is unlikely to occur to any significant level. Therefore this reaction ratio is assumed as purely a general and theoretical estimate of O2 consumption by SO2.

11.2.3. Oxygen uptake

Given oxygen exposure, the oxygen uptake of wine is around 1-2 mg/l/day. This is a generalisation however, since different wines have different rates of O2 consumption, and this appears to change with increased O2 exposure (Rossi and Singleton 1966, Perscheid and Zürn 1977).

It takes several days for O2 saturated wine (8-8.6 mg/l O2) to be consumed by SO2 (in a synthetic medium). At room temperature, dissolved O2 usually drops to undetectable levels after about a week. SO2 depletion
Over time, free SO2 decreases in wine. Average estimates indicate that SO2 depletion may be around 5 mg/l per month in wines stored in large tanks in cool cellars with small headspaces. Wines stored in warm cellars with large headspaces often lose 10-20 mg/l per month, or more. [Eisenman, 2001]. (SO2 in bottle exhibits a depletion of no more than a few milligrams per year.) Because of this decrease, SO2 levels must be continually maintained.

SO2 depletion increases with an increase in temperature, headspace, and oxygen exposed surface area to volume ratio. Since it is dependant on many variables, SO2 depletion varies from set-up to set-up and wine to wine. Safe assumptions can be made based on the past experience witnessed with each set up. For this, free SO2 levels must have been measured to determine the level of decrease over a given time and situation. Saturation level
The saturation level of dissolved oxygen in juice/wine depends on temperature (it increases with a decrease in temperature) and the alcohol content of the wine (it increases with an increase in alcoholic content). At 20°C (68°F) 8 mg/l (6 ml/l) is the saturation level, whereas at 0°C (32°F) it is 11 mg/l (8 ml/l). [Peynaud, 1984, p.248] Thus, the oxygen saturation range in wine is generally 7-11 mg/l (5-8 ml/l). [Supported by Rankine, 1995, p.187-188; Jackisch, 1985, p.115]. This level cannot be surpassed unless the temperature or pressure changes.
The oxygen content of commercially extracted orange juice is reported as 2.5-4.7 ml/l [Pulley and von Loesecke, 1939], of laboratory extracted orange juice as 2.7-5 ml/l [Loeffler, 1940], and of hand-ream extracted as 5 ml/l [Kefford et al., 1950]. Though Eisenman [2001] claims oxygen saturated juice contains about 10 mg/l of oxygen.

Note that the figures quoted outside of parentheses below are in milligrams per litre and those inside parentheses are in millilitres per litre. Racking
Gentle racking often causes an oxygen uptake of 1-3 mg/l (0.8-2.3 ml/l), whereas those with more turbulence and air exposure might absorb 3-8 mg/l (2.3-6 ml/l) during each racking. [3-4 ml/l in Peynaud, 1984, p.249; 5-6 ml/l in Jackisch, 1985, p.117; 4 mg/l from Kelly and Wollan, 2003]. Barrels
Oxygen uptake by wines in barrels is highly variable. Some quote 3-7 mg/l per year (2-5 ml/l/yr). Others find 20-27 mg/l/yr (15-20 ml/l/yr) [Mountonet et al., 1998]. The highest diffusion rate has been estimated at 26.4 ml/l/yr, and oxygen exposure due to topping and ullage has been estimated at 5 ml/l/yr [Kelly and Wollan, 2003]. This increases with less close-grained wood and smaller cask sizes - in tuns of 5cm thickness it was considered to be practically nil [Peynaud, 1984, p.248].

However, when considering that barrels are often opened for testing/tasting, oxygen absorption may be around 40-53 mg/l per year (30-40 ml/l/yr) [Jackisch, 1985, p.117]. Peynaud notes that absorption through surface exposure is about 20-27 mg/l per year (15-20 ml/l/yr), whether at the bung-on-top position with regular toppings or the bung-on-side position [Peynaud, 1984, p.248]. A partially filled container of wine with a surface area of 100 cm2 will absorb oxygen at 2 mg/l per hour (1.5 ml/l/hr). [Peynaud, 1984, p.248]

(The conversions from oxygen's volumetric measures to oxygen's by weight measures are calculated at 1 atm and 20°C (68°F). Under these conditions, 1 ml/l of oxygen weighs 1.33 mg/l. At 0°C (32°F) and 1 atm it's 1.43 mg/l, a difference of only 7% which is considered reasonably comparable.)

11.2.4. pH alteration due to sulphate formation

Since wine is an acid solution, H+ ions are present and the sulphate (2SO42-) forms sulphuric acid (HSO4-). (See reaction equations below.) The formation of sulphuric acid lowers pH. This can result in a harsher tasting wine. However, the production of sulphuric acid is small (0.82 g/l titratable acidity as tartaric acid when 350 mg/l SO2 is used). In wines with botrytised fruit, and non-botrytised sweet wines with high SO2 concentrations, a considerable amount of sulphate can be formed (0.5 g/l as tartaric). For wines stored in barrels over long periods, this can result in reduced wine quality.

SO2 + H2O ===> HSO3- + H+
2HSO3- + O2 ===> 2SO4>= + 2H+

11.2.5. Overview

Given that SO2 does not react directly with O2, all measures to protect wines from O2 should be taken to avoid O2 exposure when the desire is to minimise wine oxidation. The presence of SO2 does not guarantee the avoidance of oxidation entirely.

12. Hyperoxidation

12.1. Hyperoxidation theory

Members of "the brown juice club" do not add SO2 to white wines before fermentation. The intention is to allow rapid polyphenol oxidase (PPO) enzymatic oxidation of the many phenolic compounds in the juice which would later be chemically oxidised in the wine. The brown quinone polymers formed are adsorbed to solids and precipitate during or soon after fermentation. The process of adding oxygen to musts to achieve this result is called "hyperoxidation" or "hyperoxygenation".

Wines made in this way are claimed to be more stable with regard to later SO2 additions and less susceptible to oxidation later in their life. The technique enhances colour stability [Müller-Späth, 1977]. It may have a desirable impact on aroma (Chardonnay) [Müller-Späth, 1988; Fabre, 1998; Cheynier et al., 1989], however it can reduce the aromatic intensity of varieties (for e.g., Sauvignon Blanc) [Dubourdieu and Lavigne, 1990]. Indeed, some (e.g. the UC Davis team) believe this technique reduces varietal character and is disadvantageous for making fruit-driven wine styles in particular. For example, the volatile sulphur compounds (4-MMP, 4-MMPOH and 3-MH) contribute significantly to the characteristic varietal aroma of Sauvignon Blanc (smelling of box tree, citrus, and grapefruit/passion fruit, respectively) [Denis Dubourdieu, 2004 noted in Zoecklein [2005]]. They are easily oxidised. Therefore, to retain varietal character it is important to protect Sauvignon Blanc juice from oxidation.

The success of hyperoxidation depends on the amount of oxygen required to fully oxidise the flavonoid phenols in the juice, the pH and temperature, and the fraction of phenols which will oxidise under PPO activity. Inconsistent results have been found with the use of this technique because of these variables. Firstly, the PPO activity of juices varies widely (even within juices of the same variety) [Traverso-Rueda and Singleton, 1973; Hooper et al., 1985]. This means that some juices will require more exposure to oxygen than others. For example, Perscheid and Zurn [1977] found that 40 oxygen saturations were required before any significant decrease in the rate of PPO activity was observed. Whilst Amano et al. [1979] saw significant decreases as soon as the juice was exposed to oxidative treatment.

The premise behind this technique is that the phenols oxidised by PPO enzymes are the same as those which will later be oxidised and browned in the wine. This may not be the case and, in fact, many flavonoides are not oxidised fully through PPO activity [Singleton, 1987]. Additionally, the lack of SO2 in must can contribute to the occurrance of a slow or stuck fermentation [Zoecklein, 2001, under "Sulfur Dioxide in the Fermenter"]. The consumption of significant amounts of oxygen in the must can potentially lead to an insufficient oxygen content for healthy yeast growth.

12.2. Practical aspects of hyperoxidation

To successfully manage this technique, PPO activity should be maximised. Aside from avoiding SO2, this also means the juice should not be fined or clarified in any way. The juice should be sparged with pure oxygen (or else compressed air) whilst mixing the juice significantly. Around 20-30 mg/l oxygen (15-23 ml/l oxygen) or 95-140 mg/l air (70-105 ml/l air) is required. Repeating the sparging procedure is recommended. The juice should then be separated (clarified) from the precipitated oxidised phenols. SO2 may be added after clarification, but is usually avoided altogether.

13. Accounting for SO2 Binding: Practical Examples

13.1. Approximations for SO2 binding

Having made an SO2 addition, winemakers should re-test the free SO2 concentration some days after the addition has been made to assure that the free SO2 level in a wine has been attained. Of course, it is preferable to make a single addition which will, having accounted for the SO2 binding which will occur, arrive reasonably close to the desired level. Using the approximate rules of SO2 binding outlined in Section 5.5 above, the following examples detail how winemakers might account for such binding.

13.2. Accounting for binding lost to bisulphite addition products

Using the rule that 50% of any SO2 addition becomes bound whilst the total SO2 content of the wine is under 50 mg/l, and that 10% of any addition becomes bound thereafter, the following example illustrates the additions a winemaker might make.

35 mg/l of SO2 is added to a white must at crush. Following fermentation, the wine has a pH of 3.1 and it is (safely assumed or) assessed that the free SO2 content is negligible and all SO2 is bound. It is desired to take the molecular SO2 level to 0.6 mg/l. 12 mg/l free SO2 is required for 0.6 mg/l molecular at pH 3.1 (see Figures 6 and 7 or Table 2 above). If all SO2 added became free, 12 mg/l would be added to obtain this level. However, it has been assumed (from the above rule) that 50% of the SO2 addition will become bound. Thus, for 12 mg/l to remain after binding, 24 mg/l (12*2 or 12*100/50) SO2 must be added.

Some time later when the wine is bulk ageing, the total amount of SO2 that has been added to the wine is larger than 50 mg/l. The pH remains 3.1 and the free SO2 has depleted to 10 mg/l.
Again, 12 mg/l free SO2 is required for 0.6 mg/l molecular at pH 3.1 (see Figures 5 and 7 or Table 2 above). Since 10 mg/l is already present, 2 mg/l (12-10) free SO2 is therefore the required addition assuming no binding occurs.

If, instead, it is assumed that binding still occurs at the lower rate of 10%, then 10% of all SO2 added at this stage becomes bound. 2.2 mg/l (2 / 90% which is also 2 / (90/100)) SO2 is required to be added to the wine to obtain the 12 mg/l free SO2 level for 0.6 mg/l molecular SO2.

13.3. Accounting for Oxygen Binding: Examples

The bottling of 5 litres of wine is conducted with some splashing. It was assumed that the wine would become almost saturated with oxygen after such a racking and 12 ml of headspace would remain in the bottle once corked. Two calculations are required: (a) to account for the SO2 required to bind with the oxygen uptake during the bottling operation, and (b) to account for the SO2 required to bind with the airspace in the bottle. It is assumed that SO2 binding with wine compounds is negligible in this case (which is likely by the time bottling is due).

(a) 7 mg/l of oxygen is assumed to be dissolved into the wine following the racking procedure. A maximum of 14 mg/l of free SO2 (7*2) is required. 70 mg might be added to the bulk 5 litres (14*5), or alternatively, 10.5 mg to each 750 ml bottle (14*0.75).
(b) Each bottle contains 12 ml of airspace. Using the fact that air is 21% oxygen, the oxygen content in the headspace is 2.5 ml (12*0.21). 2.5 ml weighs 3.3 mg (2.5 mg * 1.33 mg/ml). Therefore 6.6 mg of SO2 is required (3.3*2) to bind with the oxygen in the headspace in each bottle.

If the SO2 is added to each bottle and not to the bulk, the total amount of SO2 in each bottle should be 17.1 mg (10.5+6.6).

13.4. Accounting for Oxygen and Binding: Combined Example

Combining the above sections, a typical SO2 addition accounting for both bound SO2 (due to wine components) and oxygen binding (using up the oxygen the wine is exposed to during, for example, a racking) may be calculated.

A wine is to be racked gently. The current free SO2 level is assumed to be zero, and the total SO2 level is under 50 mg/l which means that approximately 50% of all added SO2 will become bound. The oxygen uptake due to racking is assumed to be 3 mg/l. The pH is 3.1 and the aim is to obtain 0.8 mg/l molecular SO2.
Accounting for the racking oxygen uptake, 6 mg/l SO2 is required (3*2). For 0.8 mg/l molecular at pH 3.1, 16 mg/l is required. Thus, a total of 22 mg/l is required assuming no wine component binding. Yet 50% of the SO2 amount added will become bound, so 44 mg/l is required (100/50*22).

14. Testing for SO2 (Ripper and AO methods)

Due to its widespread and historic use in the wine and the food industries in general a number of analytical methods exist for measuring SO2. The most common methods are probably the Ripper method [Ripper, 1892] and the aeration oxidation (AO) method.

14.1. Ripper method

The Ripper method for SO2 uses an iodine standard to titrate the SO2 in a sample. Free SO2 is determined directly while total SO2 can be ascertained by treating the sample with sodium hydroxide before the titration to release bound SO2.

The free and total SO2 analysis used in the Ripper test is based on the redox reaction:

SO2 + 2H2O + I2 --> H2SO4 + 2HI

A starch indicator is added to the wine sample and it is acidified with H2SO4. The sample is then rapidly titrated with an iodine solution. The completion of the reaction is noted when excess iodine is complexed. This is determined by a blue-black colour end point in the presence of a previously added starch indicator.

A simple and cheap way to conduct a Ripper test is to use Chemetrics "Titrets" kits ( (See Figure 10.)

Figure 10. Using a Titrets ampoule to verify the SO2 content of a stock solution.
The Ripper method is, however, slightly inaccurate. The method suffers from the fact that the iodine reacts with oxidisable substances in wine (e.g. phenols, ascorbic acid), resulting in increased consumption of iodine and subsequently a false-high SO2 estimation. The correct assessment of free SO2 using this method is susceptible to the fact that bound SO2 interferes with the measurement. The reduction of the free SO2 during both the Ripper and AO methods results in a low bisulphite level and consequently, some bisulphite is released from the bound SO2 to re-attain equilibrium. In red wines, the release of SO2 from bound SO2-anthocyanin can significantly result in a false high measurement of free SO2. How rapidly the carbonyl bound SO2 compounds dissociate to release free SO2 depends on the dissociation rates for SO2 binding with those respective compounds (see section 5.3. above). Acetaldehyde bound SO2, for example, will be slow to release. However, pyruvate (and possibly alpha-keto-glutarate) will dissociate faster. Wines high in pyruvate will therefore result in an increased release of free SO2 from the bound form, yielding a false-high measurement of the free SO2 content of the wine. This is another reason why such titrations should be conducted rapidly.

It can also be particularly difficult using this method with reds, since the dark colour of red wines makes it difficult to identify the end point of the titration. Use of an oxidation-reduction electrode will not solve this problem since the end point is dependent on the actual blue starch end point.

Additionally, the potential volatilisation of SO2 during titration, and the reduction of the iodine titrant by non-sulphite compounds such as phenols or pigments, can effect the result significantly. Other interferences include botrytis and ascorbic acid (results are false-high due to the competitive oxidation of ascorbic acid and SO2 by the iodine titrant).

Despite the inaccuracies, the Ripper method remains the most common method used for free SO2 determination in winemaking (including commercial winery labs) due to its speed and simplicity.

14.2. AO method

The AO method, also called the Rankine method after Rankine [1962] or the Tanner method after Tanner [1963], is a modification of the Monier-Williams method. It involves the acidification of a sample, followed by the distillation of SO2 (with nitrogen sweeping gas or air aspiration) out of the sample into a peroxide solution. The SO2 and peroxide react to form H2SO4 according to the following equation:

H2O2 + SO2 --> SO3 + H2O --> H2SO4

The acid formed is then titrated with NaOH to an end point, and the volume of NaOH required used to calculate the SO2 level. This method avoids the iodine-phenol binding which occurs in the Ripper method. However, it is not without its problems. The acidification of the solution causes a shift in the equilibrium between bisulphite and anthocyanins, and the SO2 bound anthocyanins (red wine pigments). This causes a freeing of some anthocyanin bound SO2, resulting in an increased free SO2 concentration and a false-high free SO2 measurement. An inefficient condenser (used in the distillation) accompanied by high volatile acidity will also result in a false-high measurement. Additionally, the flow rate of the aspiration is important in AO testing. An excessively high aspiration rate may prevent sufficient time to allow the H2O2-SO2 reaction to complete. A 15-20 minute aspiration time with a flow rate of 1-1.5 L/min is recommended [Buechsenstein and Ough, 1978].
Despite these potential inaccuracies, the AO method gives reproducible results and has been noted as having an accuracy with just a 2.5-5% error [Buechsenstein and Ough, 1978].

14.3. Countering inaccuracies

To counteract the inaccuracies due to these testing techniques, winemakers sometimes dilute the sample with distilled water. This helps determine the end-point but lowers accuracy. Additionally, dilution may further affect the free-total SO2 equilibrium.
The use of hydrogen peroxide (H2O2) may also be employed to obtain better free SO2 evaluations. In this case, a wine sample is titrated in the usual way. Then a second sample is treated with excess H2O2 to remove the free SO2, and is subsequently titrated. The free SO2 in the wine is then assumed to be the value obtained in the first titration minus the value obtained in the second titration. The argument here is that the H2O2 removes all the SO2 present such that, when the sample is then titrated, the iodine is reacting with the phenols in the sample. Thus, the value obtained when solely accounting for phenol-binders is subtracted from that when they are included, leaving the true SO2 value. It is possible that the addition of the H2O2 in this method upsets the equilibrium between free and total SO2 leading to a false reading. Nevertheless, numerous winemakers have found this method compares favourably with values obtained from the AO method [e.g. Eisenmann, 2004].
The use of a high intensity light to illuminate the sample (especially with reds), and the use of calibrated control solutions may also help.

Use of a baseline value might also improve the accuracy of testing results. This is achieved by testing for free SO2 immediately after fermentation (before any SO2 addition has been made), when the free SO2 content in the wine is expected to be zero. The testing result obtained at this time is then used to correct all later test result values. Of course, this method will fail to give accurate results when the SO2 content in newly fermented wine is not zero.

14.4. Applying value adjustments

Value adjustments ("fudge factors") are usually applied to the readings to account for their inaccuracies. Such value adjustments are based on previous values attained from more reliable sources. For example, value adjustments made to values obtained from the Ripper method might be corrected based on values made by historical comparison using the alternative aeration oxidation (AO) method.

The true SO2 content obtained using the Ripper method is often considered to be an overestimate of the true value by around 10-20 mg/l (some quote 10 for whites and 20 for reds, whilst others do not adjust for whites at all). Breeden [2002a] noted that the Ripper method commonly measured 15-20 mg/l false high for red wines and 8 mg/l high for Chardonnay wines when compared with measured values using the AO method. Additionally, he claimed differences of up to 30 mg/l were possible [Breeden, 2002b]. He suggested a correction factor (subtraction from Ripper result value) of 20 mg/l be used on reds [Breeden, 2003]. Results obtained by Eisenman [2004] comparing Ripper to AO show a similar disparity. He used a correction factor of 15-20 mg/l for medium-heavy reds, respectively. The largest difference between AO and standard Ripper that he had encountered was 19 mg/l (with a dark, tannic Syrah). Breeden also compared results obtained from Titrets and regular Ripper. Testing showed that, on average, Titrets overestimated regular Ripper values by an average of 6 mg/l (differences ranged 3-8.5) for Chardonnay (and one Pinot Noir) wines [data analysed from Breeden, 1999]. Such differences may be deemed due to operator error only.

This data suggests that the difference between values obtained from AO and Ripper methods varies from zero to 20 or 30 mg/l, depending on the wine style. For less phenolic wines, the difference may be typically zero to 10 mg/l, whilst for more heavily extracted wines the difference may be 15-20 mg/l. These factors may be used when correcting values obtained with the Ripper method. Of course, such correction factors assume that the AO method gives completely accurate measurement of SO2 in wine, which is not precisely the case.

14.5. Minimising operator error

With any analytical test method there is the risk of operator error entering into the determination. Aside from the usual issues of correct reagant concentrations, it should be borne in mind that SO2 is a volatile substance. Samples of wine used for SO2 analysis should be tested as soon as possible after taking them from the bulk wine. In cases where this is not possible, air contact with the sample should be minimised. Samples should be taken from a homogeneous source. If the wine sample for testing is solely sourced from near the headspace of the storage vessel or, alternatively near/amongst the lees, the SO2 concentration is likely to be slightly lower than elsewhere. Samples should be made with this in mind, and should be taken to be representative of the vessel as a whole.

15. Removing Free SO2

Sometimes excessive SO2 is (accidentally) added to a wine. SO2 will slowly deplete during an oxidative ageing process, but sometimes winemakers wish to reduce the SO2 level in a short period of time. In any case, desulphiting may inevitably result in loss of aroma/flavour [Weger, 1956] and overdosing should be avoided at all costs. There are three methods commonly employed for such situations, and a fourth potential method.

15.1. Blending

Blending the high-SO2 wine with another wine low in SO2 is the safest method, ensuring the wine does not suffer from oxidation or further processing.

15.2. Aeration

SO2 is often removed from wine by aerating. This is based on the slow oxidation of the SO2 and is only really suitable for slightly excessive doses of SO2, since excessively high doses will not be successfully stripped by even multiple aerations.
Usually the wine is transferred from one vessel to another in a violent manner (with turbulence) to encourage oxygen contact. This method can be traumatic for a wine, potentially over-oxidising and "damaging" its delicacy. However, it remains a simple solution to reducing excessive SO2. A wine saturated with oxygen will contain 5-8 mg/l oxygen (see section "SO2 and Oxidation, Saturation level" above). Assuming a complete reaction (though somewhat chemically unrealistic), this amount of oxygen may remove 20-32 mg/l SO2. If the aim is to reduce SO2 by over 20-32 mg/l then this method can be used on a periodic basis more than once (with several days between successive operations). If the aim is to reduce the SO2 by less that 20 mg/l, the aerating should be done with less violence.

15.3. Hydrogen Peroxide

15.3.1. Theory

Free SO2 can be removed by adding hydrogen peroxide (H2O2) to wine. The use of H2O2 is considered too severe by many. Nevertheless, it remains one of the only real options for removing excessively high levels of SO2 from wine for the non-commercial winemaker.

The removal reaction is:
SO2 + H2O2 ===> SO4--+ 2H+

The molecular weight of SO2 is 64.1 and that of H2O2 is 34. Therefore, 0.5304 g (1/64.1*34) of H2O2 is required to react with 1 g of SO2.

The peroxide reacts with molecular SO2, changing the SO2 equilibrium. Since this equilibrium is continually re-establishing, the H2O2 should be added slowly. Additionally, since H2O2 is such a powerful oxidiser, the amount added should be calculated carefully. Analytically testing the SO2 content before and after H2O2 addition is advised.

Solutions of H2O2 commonly come as 3% solutions. If they are mass/mass solutions (this appears to be the typical case) they should contain about 30.3 mg/ml H2O2. If they are volume/volume solutions they should contain about 42.3 mg/ml H2O2. (See "Information on H2O2 content" below for more details.)

15.3.2. Example using H2O2

15 litres of wine has a free SO2 level of 70 mg/l. It is desired to reduce this to 40 mg/l. The reduction of 30 mg/l (70-40) requires an H2O2 addition of 16 mg/l (0.5304*30). Thus, the 15 litres requires an addition of 240 mg (15*16) of H2O2. Using a 3% mass/mass solution of H2O2, 7.9 ml (240/30.3) of the solution needs to be added to the 15 litres for the drop to 40 mg/l.

15.3.3. Information on H2O2 content

Pure (100%/weight) H2O2 has a density of about 1.41 g/ml.
Mass/mass solutions: 3 g H2O2 / (97 g H2O + 3 g H2O2) means a volume of 97 ml + (3 g / 1.41 g/ml = 2.13 ml) = 99.1 ml. This contains 3 g per 99.1 ml which is 30.3 mg H2O2/ml of the 3% solution.
Volume/volume solutions: 3 ml H2O2 / (97 ml H2O + 3 ml H2O2). 3 ml H2O2 provides (3 ml * 1.41 g/ml =) 4.23 g H2O2 per 100 ml solution, which is 42.3 mg H2O2/ml of the 3% solution.

15.4. Inert Gas Stripping

This technique is used to remove SO2 from large-scale commercial fruit juices. On a small scale, it might be done by sparging a receiving vessel with CO2 (or nitrogen or argon). The wine is then sprayed against the vessel wall in an attempt to volatise the SO2. Alternatively, bubbling inert gas through the wine might be practised. The effectiveness of this method is, however, questionable without the use of sophisticated equipment.

16. Adding SO2: Practical Considerations

When adding SO2, it is important to ensure that it is evenly distributed in the must or wine. Injecting SO2 solution steadily (rather than in a single hit) during pumping/transfer/racking procedures presents an ideal method of homogenous SO2 addition.

Figure 11. Localised discolouration of pomace indicating high point-concentration of SO2.
Due to the rapid enzymatic oxidation reactions in musts, SO2 ought to be in contact with the juice as soon as possible after crushing the fruit. This is the principal which should be followed in any SO2 additions to musts. Exactly how this is practised may vary from set-up to set-up. In the case where fruit is partially crushed upon harvesting, SO2 should be added to the fruit with the aim to take action within the juice resulting from partial crushing.
Addition of SO2 to the uncrushed fruit in the case of reds, or crushed and unpressed fruit in the case of whites, will result in SO2 binding with fruit solids. Such binding should be accounted for, and higher SO2 additions may be required in such situations.
Oxidation of draining press juice will be significant in the absence of SO2 and SO2 might therefore be added to the marc of the post-free run press fraction. Delteil [2001] argues that this practise results in increased aromatics and varietal expression, greater palate volume and decreased sensations of palate dryness.

Post crush additions in a liquid form are recommended, since they assist in SO2 distribution and help prevent combination with solids. SO2 is most effective when added to individual portions of the must during, or within quick succession of, pressing (or crushing, in the case of reds). This method presents a more effective use of SO2 than a number of consecutive additions to must storage vessels (e.g. must receptacle tanks/vessels).

High point-concentrations of SO2 indicate that SO2 has not been mixed thoroughly. In practice, this can sometimes be seen in crushed fruit or fresh must as a localised discolouration of the pomace or juice. Figure 11 shows this phenomenon.

According to some, adding small concentrations of SO2 to must sequentially results in more oxidation than would occur in an unsulphited juice. Additionally, SO2 doses are more effective on yeast and microbes if the dose is given as a single high dose, rather than a number of small sequential doses.

Since a portion of any previously added SO2 is in the bound form and therefore not effective, SO2 solutions used for additions should be relatively dilute.
After fermentation, corrective SO2 additions should only be made under conditions of potential contamination or volatilisation (e.g. during transfer or under high temperatures) or when molecular SO2 levels are far from the required effective dose.

17. Typical SO2 Additions

Winemakers who add SO2 pre-fermentation typically add around 25-50 mg/l at crush. This is followed by a post alcoholic fermentation (or post malolactic fermentation) addition sufficient to, having accounted for binding, maintain the desired molecular SO2 level. (Some estimate this as 120-150% of the amount required to maintain the desired molecular SO2 level.) During bulk ageing, and for bottling, the wine is maintained at this same molecular SO2 level.

As mentioned previously, molecular SO2 levels are pH dependent. However, many winemakers cannot assess pH in their wines and, therefore, quantities of total SO2 to add at particular times or procedures of winemaking are made based on rough guidelines. Exact quantities vary from winemaker to winemaker (and on wine type/style and set-up). However, dosages can be amplified or reduced depending on the circumstances surrounding the quality of the fruit, juice, and wine. For fruit and musts, the following situations will require increased SO2 dosages: high suspended solids, ruptured/diseased fruit, violent or prolonged fruit transport, increased handling, higher temperatures, or a longer duration between crush and fermentation. For wines, increased temperatures and increased exposure to air tend to call for increased SO2 dosages. At bottling, wine style and the intended duration of ageing dominate dosage decisions.

It is worth noting that, because of the differences in environmental conditions and typical practises in different countries, typical additions vary among different countries and regions. The additions in France, for example, are often much higher than what is considered necessary or normal in California. Likewise, hotter climates tend to receive higher doses (e.g. Languedoc Roussillon vs Burgundy). Common SO2 levels for addition to must are presented in Table 4 and levels for addition to wines are presented in Table 5. Note that these values are not the dosage additions themselves, but are the quantity of free SO2 that should exist in the must/wine after addition (binding should be taken into account upon addition to ensure that these levels are met). Table 6 shows recommended maximum values of total SO2. On an international scale, these values are relatively conservative.

(For typical Campden tablet additions, see the Campden Tablets section below.)

Figure 12. Weighing metabisulphite powder.

Table 4. Recommended free SO2 levels for musts
Circumstance Free SO2 (mg/l)
white, healthy fruit, low pH 25-50
white, healthy fruit, high pH 60-80
white, fruit with some rot 80-100
red, healthy fruit, low pH 50
red, healthy fruit, high pH 50-80
red, fruit with some rot 80-100

Table 5. Recommended free SO2 levels for wine
Circumstance Free SO2 (mg/l)
before MLF none / under 20
dry white, maintenance 30-40
sweet white, maintenance 40-80
red, maintenance 20-40
dry white, bottling 20-30
sweet white, bottling 30-50
red, bottling 10-30

The range in the values corresponds to the pH of the wine. If the wine is an acidic style the lower values should be used, whilst the higher values should be used for less acidic wines. Wines which will be travelling or stored in unfavourable conditions often have 1.5 to 2 times the bottling values above (Table 5) added at bottling.

Table 6. Recommended total SO2 levels for wine
white (conservative) under 150
red (conservative) under 150
white, dry (liberal) under 200
white, white (liberal) under 400
red (liberal) under 300

It is sometimes claimed that SO2 additions without reference to pH is sufficient. Whilst this is generally true, there are exceptions. Figure 13 shows three different levels of free SO2 (20, 30, and 50 mg/l) and their corresponding molecular SO2 levels at varying pHs. Assuming that molecular SO2 must be kept below the sensory threshold (2 mg/l) and above 0.8 mg/l for microbial stability (see Section 6), then "safe zones" are molecular SO2 levels between these two values (i.e. 0.8-2 mg/l). It can be seen from the figure that these zones still vary over the typical pH values encountered in wine. For example, whilst 30 mg/l free SO2 is sufficient for pH values ranging 3.0 to 3.4, it is not suitable for pHs outside this range (below this level it is above sensory threshold, above this level it is below levels suitable for microbial stability). Based on this information, it might be suggested that red wines (which typically have a pH > 3.2) be kept around 50 mg/l free SO2 and white wines (which typically have a pH < 3.3) be kept above 30 but below 50 mg/l free SO2. Of course, awareness of the pH is always preferable to such estimates.

Figure 13. Safe zone "windows" for SO2 levels.

18. Storage and Purity

Dry sulphur dioxide (in the metabisulphite form, or otherwise) is sensitive to high temperatures and humidity. It will lose its strength under such conditions. It is important to replace SO2 stores reasonably frequently.

Additionally, the strength of SO2 is sometimes weak upon purchase. For example, upon making up an aqueous solution of metabisulphite and testing its SO2 concentration, it is not uncommon to find only 90% of the expected SO2 value present. This is most likely due to the conditions experienced by the SO2 prior to purchase. It is therefore important to check the strength of the SO2 stock being used for winemaking additions.

19. Stock Solutions

Figure 14. Stock solution and syringe.
Stock solutions of dissolved sodium or potassium metabisulphite salts provide a fast and simple way of adding sulphite to a wine. This is especially the case when a gram scale is not available and measuring a volume of stock solution is preferential to weighing out very small quantities of powder.

It is important to keep a stock solution in an air tight container since contact with air will decompose the sulphite. (It should also be noted that plastic is breathable to some extent, and stock solutions stored in plastic bottles should therefore be remade relatively frequently.)

As an example of the calculations used in making and using a stock solution, a 10% stock solution can be made up by adding enough water to 100 grams of potassium metabisulphite to make up a total volume of 1 litre (100 grams / 1000 mls * 100 = 10%). This solution contains 100 mg/ml of potassium metabisulphite. Since potassium metabisulphite is only 57.6% SO2, this solution then contains 5.76% SO2 (10% * 0.576 = 5.76%) or, alternatively stated, it contains 57.6 mg/ml of SO2 (100 mg/ml * 0.576 = 57.6 mg/ml).
10 ml of this 10% stock solution added to 20 litres gives 50 mg/l of potassium metabisulphite (100 mg/ml * 10 ml / 20 L = 50 mg/l) which gives 28.8 mg/l of SO2 (50 mg/l * 0.576).
Alternatively, to obtain 30 mg/l of SO2 in 15 litres, this requires 781 mg of potassium metabisulphite (30 mg/l * 15 l / 0.576 = 781 mg) for which 7.8 ml of the 10% stock solution is required (450 mg / 100 mg/ml / 57.6 % SO2 = 4.5 / 0.576 = 7.81 ml).

20. Campden Tablets

Campden tablets are designed to have a mass of 0.44 grams. However, consistency of the tablet size in manufacturing is questionable, and many winemakers claim there is little certainty that tablets contain the amount of metabisulphite they are intended to (expected concentrations have been seen to deviate by up to 25%). Additionally, some winemakers claim that the "fillers" used in Campden tablets to increase the bulk size of the tablet, taint wine flavour and affect clarity. Nevertheless, Campden tablets remain a simple way of adding a small (if rough) quantity of sulphite to a must or wine.

Rules of thumb for the use of Campden tablets are generally quoted as:
One tablet should be added per gallon (Imperial or US) initially and then one at each of the 2nd, 4th, 6th, etc rackings.
Or, if heat is used in preparing the must, none initially but one per gallon at each of the 1st, 3rd, 5th, etc rackings.

Assuming one Campden tablet contains 0.44 grams of potassium/sodium metabisulphite, the following sulphite levels are obtained by the addition of 1 tablet to the given volumes:

Table 7. SO2 Equivalent Campden Tablet Dosages
Salt per Imperial gallon per US gallon per litre
Sodium 65 mg/l 78 mg/l 297 mg/l
Potassium 56 mg/l 67 mg/l 254 mg/l

In practice, these figures may vary by up to 25%, possibly more.

Figure 15. Campden tablets - no need to weigh.

21. Sulphur Wicks and Rings

Sulphur wicks or rings are usually comprised of cellulose coated sulphur or a mineral (aluminium or calcium silicate) mixed with sulphur. They are generally only used for dosing barrels or small wooden tanks. They are not recommended for use in concrete tanks or stainless steel, due to the subsequent chemical attack on the surfaces of these vessels.

Sulphur wicks and rings are heterogeneous and their exact SO2 content varies due to the manufacturing process and storage conditions (again, humidity reduces their effectiveness) [Chatonnet et al., 1993]. Sulphur rings are more sensitive to storage conditions than wicks.

In theory, the burning of a sulphur wick or ring follows the reaction:

S + O2 ===> SO2

However, in reality around 20-30% of the sulphur is lost due to (1) part of the sulphur falling from the wick before it burns and (2) part of the sulphur producing sulphuric acid. In an enclosed space like a barrel, the amount of sulphur which can be burnt is limited by the presence of sulphurous gas which inhibits combustion. For example, in a 225 litre barrel around 20 g of sulphur is the limit which can be burnt. In addition to this, humid barrels hinder combustion. In general, 5 g of sulphur burnt in a 225 litre barrel will increase the SO2 by 10-20 mg/l (by burning a sulphur ring) or by 10 mg/l (by burning a sulphur wick) [Chatonnet et al., 1993].

SO2 is not usually distributed evenly during the filling of sulphured barrels, nor does it homogenise well afterwards. A thorough mixing is therefore recommended after barrel filling (e.g. by rolling the barrel).

22. Acknowledgements

The author would like to thank Dr. John Danilewicz for his assistance in accurately describing the reactions regarding the oxidation of wine in the presence of sulphur dioxide. There is much debate with regards to the theories expounded in the literature in this area and Dr. Danilewicz's cutting edge input in this area is much appreciated.
Additionally, the author would like to thank Lum Eisenman for his communication of AO and Ripper test data results.

23. References

Amano, Y., Kubota, M., and Kagami, M. (1979). Oxygen uptake of Koshu grape must and its control. Hokkokogaku Kaishi 57:92-101.
Amerine, M.A. and Joslyn, M.A. (1951). Table wines; the tech. of their prod. Berkley and LA, Univ. Cal. Press.
Beech, F.W., Burroughs, L.F., Timberlake, C.F., and Whiting, G.C. (1979). Progres recents sur l'aspect chimique et antimicrobienne de l'anhydride sulfureux. Bulletin OIV 52(586):1001-1022.
Berg, H.W. and Akiyoshi, M. (1962). Color behavior during fermentation and aging of wines. Am. J. Enol.Vitic. 13:126-132.
Berg, H.W., Filipello, F., Hinreiner, E., and Webb, A.D. (1955). Evaluation of threshold and minimum difference concentrations for various constiuents of wines: I. Water solutions of pure substances. Food Tech. 9:23-26.
Biedermann, W. (1956). Oxidation in fruits and fruit juices. Mitt. Lebensmittelunters. Hyg. 47, 86-112. (German).
Bioletti, F.T. (1912). Sulfurous acid in winemaking. 8th Int. Cong. Appl. Chem. 14:31-59.
Blouin, J. (1963). Constituants du vin combinant de l'acide sulfureux. Ann. Technol. Agr. 12 (numéro hors série 1):97-98.
Blouin, J (1966). Contribution a l'edute des combinaytions de l'anhydride sulfureux dans les mouts et les vins. Ann. Technol. Agr. 25:223-287, 360-401.
Braverman, J.B.S. (1963). Introduction to the Biochemistry of Foods. Amsterdam: Elsevier Publishing Co.
Breeden, D.C. (1999). Comparison of Chemetrics Titrets to Regular Ripper, rec.crafts.winemaking Usenet newsgroup thread, 23/06/1999.
Breeden, D.C. (2002a). SO2 Titrette Accuracy. rec.crafts.winemaking Usenet newsgroup thread, 2002-03-14.
Breeden, D.C. (2002b). too much sulfite. rec.crafts.winemaking Usenet newsgroup thread, 2002-02-10.
Breeden, D.C. (2003). SO2 in wine. rec.crafts.winemaking Usenet newsgroup thread, 2003-08-26.
Buechsenstein, J.W. and Ough, C.S. (1978). SO2 determination by aeration-oxidation: a comparison with Ripper. Am. J. Enol. Vitic. 29, 161.
Burroughs, L.F. (1975). Determining free sulfur dioxide in red wines. Am. J. Enol. Vitic. 26:25-29.
Burroughs, L.F. and Sparks, A.H. (1973a). Sulphite-binding power of wines and ciders. I. Equilibrium constants for the dissociation of carbonyl bisulphite compounds. J. Sci. Food Agric. 24:187-198.
Burroughs, L. and Sparks, A.H. (1973b). Sulphite-binding power of wines and ciders. I. Equilibrium constants for the dissociation of carbonyl bisulphite compounds. J. Sci. Food Agric. 24:187-198.
Burroughs, L. and Sparks, A.H. (1973c). Sulphite-binding power of wines and ciders. III. Determination of carbonyl compounds in a wine and calculation of its sulphite-binding power. J. Sci. Food Agric. 24:207-217.
Burroughs, L. and Whiting, G. 1960. The sulphur dioxide combining power of cider. Ann. Rept. Agr. Hort. Exper. Stat. Long Ashton 1960:144-147.
Chatonnet P., Boidron, J.N, and Dubourdieu, D. (1993). J. Int. Sci. Vigne Vin, 27 (4), 277-298.
Cheynier, V., Rigaud, J., Souquet, J.M., Duprat, F. and Moutounet, M. (1989). Am. J. Enol. Vitic. 40:1, 36-42.
Cruess, W.V. (1912.) The effect of sulfurous acid on fermentation organisms. J. Ind. Eng. Chem. 4:581-585.
Danilewicz, J.C. (2003). Review of Reaction Mechanisms of Oxygen and Proposed Intermediate Reduction Products in Wine: Central Role of Iron and Copper. Am. J. Enol. Vitic. 54(2):73-85.
Delfini, C. (1981). La stabilizzazione microbiologica in enologia mediante l'impiego dell'anidride solforosa: nuovi (o vecchi?) concetti. L'Enotecnico, 10:27-33; 11:32-33. Ann. Ist. Sper. Enologia, Asti, 12:239-250.
Delfini, C. (1984). Prove sperimentali sulla dose minima di SO2 necessaria per mutizzare un mosto. L'Enotecnico, 1:51-57.
Delfini, C. (1988). Refermentation potential in Bottled Sweet Wines of Yeasts Adpated to Sulfur Dioxide. Chem. Mikrobiol. Technol. Lebensm. 11:137-142. Delfini, C. (1989). Ability of wine malolactic bacteria to produce histamine. Sciences des Aliments, 9:413-416.
Delfini, C. (1992). Criteri metodologici seguiti, risultati ottenuti e prospetitive nella selezione di lieviti per uso enologico. Biologia Oggi, VI (1-2):303-309.
Delfini, C. and Morsiani, M.G. (1992). Study on the resistance to sulfur dioxide of malolactic strains of Lueconostoc oenos and Lactobaccillus sp. isolated from wines. Science des Aliments, 12:493-511.
Delteil, D. (2001). Mastering SO2 during the pre-fermentation phases of white wine-making. ICV Flashinfo, available at, September.
Dittrich, H.H. and Staudenmayer, T. (1968). SO2-Bildung, Böckserbeseitigung. D Weinztg 104:707-709.
Dott, W., Heinzel, M., Trüper, H.G. (1976). Sulfite formation by wine yeasts: I. Relationship between growth, fermentation and sulfite formation. Archives of Microbiology 107, 289-292.
du Toit, Wessel. (2000). The SO2 resistance of South African acetic acid bacteria and their effect on fermentation. Department of Viticulture and Oenology and the IWBT, University of Stellenbosch work published at
Dubernet, M. and Ribéreau-Gayon, P. (1973). Presence et significance dans les mouts et les vins de la tyrosinase du raisin. Conn. Vigne Vin 7:283-302.
Dubernet, M. and Ribéreau-Gayon, P. (1974). Vitis 13, 233.
Dubourdieu, D. and Lavigne, V. (1990). Rev. Fr. Oenol. 124, 58-61.
Eisenman, L. (2001). Oxygen Uptake in Wine. Available at The Home Winemakers Manual.
Eisenman, L. (2004). Personal communication.
Eschenbruch, R. (1974). Sulfite and sulfide formation during winemaking - a review. Am. J. Enol. and Vitic. 25 (3), 157-161.
Eschenbroch, R. and Bonish, P. (1976). The influence of pH on sulfite formation by yeasts. Arch Microbiol 107:229-231.
Fabre, S. (1998). Objectif 28, March, 21-25.
Farkas, J. (1988). Technology and Biochemistry of Wine, Vols. 1 and 2. Gordon & Breach, New York.
Fornachon. (1963). Inhibition of Certain Lactic Acid Bacteria by Free and Bound Sulfur Dioxide, J. Sci. Food Agr., 14: 857-862.
Harvalia, A. (1965). La couleur des vins rouges. Chim. Chronika (Athens) 20(9):155-159.
Heard, H. and Fleet, G.H. (1988). Austr. NZ Wine Ind. J., 3, 57-60.
Heinzel, M., Dott, W., Trüper, H.G. (1976). Störungen im Schwefelstoffwechsel als Ursache der SO2-Bildung durch Weinhefen. Wein-Wiss 31:275-286.
Hennig, K. 1943. Bilans de l'azote dans les moûts et les vins nouveaux en fermentation. Bull. O.I.V. 16(159):82-86.
Hennig, K. and Burkhardt, R. (1960a). Detection of phenolic compounds and hydroxy acids in grapes, wines, and similar beverages. Am. J. Enol. Vitic. 11:64-79.
Hennig, K. and Burkhardt, R. (1960b). Vorkommen und Nachweis von Quercitrin und Myricitrin inn Trauben und Wein. Weinberg Keller. 7:1-3.
Hood, A. (1983). Inhibition of growth of wine lactic-acid bacteria by acetaldehyde-bound sulphur dioxide. Aust. Grapegrower & Winemaker 232:34-43.
Hooper, R.L., Collins, G.G., and Rankine, B.C. (1985). Catecholase activity in Australian white grape varieties. Am. J. Enol. Vitic. 36:203-206.
Ingram, M. (1948). Germicidal effects of free and combines sulfur dioxide. J. Soc. Chem. Ind. 67:18-21.
Jackisch, Philip. (1985). Modern Winemaking, Cornell University Press, 1985.
Joslyn, M.A. and Braverman, J.B.S. (1954). The chemistry and technology of the pretreatment and preservation of fruit and vegetable products with sulfur dioxide and sulfites. In Advances in Food Research 5:97-160. New York: Academic Press.
Jurd, L. (1964). Reactions involved in sulfite bleaching of anthocyanins. J. Food. Sci. 29:16-19.
Katchmer, J. (1990). Effects of sulfur dioxide and bisulfite-binding compounds on short term yeast viability in a model wine solution. M.S. thesis, Davis, CA: University of California.
Kefford, J.F., McKenzie, H.A., and Thmopson, P.C.O. (1950). Effect of oxygen on flavor deterioration and loss of ascorbic acid in canned orange juice. Food Preservation Quarterly (Australia). 10, 3, 44-47.
Kelly, M. and Wollan, D. (2003). Micro-oxygenation of wine in barrels. The Australian & New Zealand Grapegrower & Winemaker. 473a:29-32.
Kerp, W. (1903). Ueber organisch gebundene schweflige Saure in Nahrungsmitteln. Z. Untersuch. Lebensm. 6:66-68.
Kerp, W. (1904a). Ueber die schweflige Säure im Wein. I. Allgemeines über die schweflige Säure im Wein. Arb. Gesundheitzamte 21(2):1-15.
Kerp, W. (1904b). Ueber die schweflige Säure im Wein. II. Allgemeines über die aldehydschweflige Säure im Wein. Arb. Gesundheitzamte 21(2):16-40.
Kerp, W. (1904c). Die schweflige Säure und ihre Verbindungen mit Aldehyden und Ketonen. I. Teil. Berlin, Julius Springer.
Kerp, W. and Bauer, E. (1904). Zur Kenntnis der gebundenen schwefligen Sauren. I. Arb. Gesundheitsamte. 21:180-185.
Kerp, W. and Bauer, E. (1907a). Zur Kenntnis der gebundenen schwefligen Sauren. II. 26:231-248.
Kerp, W. and Bauer, E. (1907b). Zur Kenntnis der gebundenen schwefligen Sauren. III. 26:269-279.
Kielhöfer, E. (1963). Etat et action de l'acide sulfureuax dans les vins; regles de son emploi. Ann. Technol. Agr. 12:77-89.
Kielhofer, E. and Würdig, G. (1960). Die an aldehyd gebundene Schweflige Saure im Wein. I. Acetaldehydbildung durch enzymatische und nicht enzymatische Alkohol-Oxydation. Weinberg Keller 7:16-22.
King, A.D. Jr., Ponting, J.D., Sanshuck, D.W., Jackson, R., and Mihara, K. (1981). Factors affecting death of yeast by sulfur dioxide. J. Food. Prot. 44:92-97.
Kolthoff, I.M. and Stenger, V.A. (1942). Volumetric Analysis Vol. I, Theoretical fundamentals. New York, Interscience Publishers, Inc.
Lafon-Lafourcade, S. and Peynaud, E. (1974). Sur l'action antibacterienne de l'annhydride sulfureux sous forme libre et sous forme comninée. Conn. Vigne Vin 8:187-203.
Lafourcade, S. (1952). Contribution à l'étude des activeurs et des inhibiteurs de la fermentation alcoolique des moûts des raisin. Ann. Technol. Agr. 4:5-66.
Lafourcade, S. (1955). Contribution a l'etude des activeurs et des inhibiteurs de la fermentation alcoolique des mouts de raisin. Ann. Technol. Agr. 4:5-66.
Lehmann, F.L. (1987). Secondary fermentations retarded by high levels of free sulfur dioxide. Aust. N.Z. Wine Ind. J. 2, 52-53.
Liu, J.-W.R. and Gallander, J.F. (1982). Effect of insoluble solids on the sulfur dioxide and rate of malolactic fermentation in white table wines. Am. J. Enol. and Vitic. 33, 194-197.
Loeffler, H.J. (1940). Determination of air in citrus juice. Ind. Eng. Chem., Analyt. Ed. 12, 533-534.
Lu Valle, J.E. (1952). The reaction of quinone and sulfite. I. Intermediates. J. Am. Chem. Soc. 74:2970-2977.
Lüthi, H. (1953). Recent analysis of the significance and preservation of oxidative reactions in apple juice. Int. Fruchtsaft-Union. Zeg. (German).
Lüthi, H. (1954). The significance of the oxidation of our fruit juices by enzymes. Schweiz. Ztschr. Obst- und Weinbau, 63, 455, 469, 494. (German).
Lüthi, H. (1960). The determination of some qualitative factors in alcohol-free fruit juices . Flüssges Obst. 25 (4), 16-19. (German).
Macris, B.J. and Markakis, P. (1974). Transport and toxicity of sulphur dioxide in Saccharomyces cerevisiae var. ellipsoideus. J. Sci. Fd. Agric. 25,21.
Margalit, Yair. (1990). Winery Technology and Operations, San Francisco: Wine Appreciation Guild Ltd., 1990.
Margalit, Yair. (1996). Concepts in Wine Chemistry, San Francisco: Wine Appreciation Guild Ltd., 1996.
Mayer, K., Vetsch, U., and Pause, G. (1975). Hemmung des biologischen Saurabbaus durch gebundene schweflige Saure. Schw. Z. Obst-Wein. 23:590-596.
Minarik, E. (1978). Progres recents dans la connaisance des phenomenes microbiologiques en vinification. Bull. O.I.V. 51(567):352-367.
Mountonet, M., Mazauric, J.P., Saint-Pierre, B., Hanocq, J.F. (1998). Gaseous Exchange in Wines Stored in Barrels. J. Sci. Tech. Tonnellerie, 4, 131-145.
Müller-Späth, H. (1977). Die Weinwirtschaft, 6, 1-12.
Müller-Späth, H. (1982). Die Rolle der Kohlensaure beim Stillwein. Weinwirt. 118:1031-1037.
Müller-Späth, H. (1988). Objectif 28, March 21-25, 15-19.
Fabre, S. (1998). Objectif 28, March, 21-25.
Osborne, J.P., Mira de Orduña, R., Pilone, G.J., and Liu, S.-Q. (2000). Acetaldehyde metabolism by wine lactic acid bacteria. FEMS Microbiology Letters 191, 51-55.
Ough, C.S. (1959). Personal communication in Amerine et al. (1951).
Ough, C.S. (1985). Some effects of temperature and SO2 on wine during simulated transport or storage. Am. J. Enol. Vitic. 36:18-22.
Ough, C.S. and Crowell, E.A. (1987). Use of sulfur dioxide in winemaking. J. Food Science 52(2):386-89.
Perscheid, M. and Zurn, F. (1977). Der einfluss von Oxydationsvorgangen auf die Weinqualitat. Weinwirt, 113:10-12.
Peynaud, Emile. (1984). Knowing and Making Wine, english translation 1984, John Wiley & Sons.
Peynaud, E. and Lafon-Lafourcade, S. 1966. Facteurs de la formation des acides pyruvique et alpha-cétoglutarique au cours de la fermentation alcoolique; conséquences pratiques sur les combinaisons sulfitiques des vins. Ind. Aliment. Agr. (Paris) 83:119-126.
Porchet, B. (1931). Contribution a l'etude de l'adaptation des levures a l'acide sulfureux. Ann. Agr. Suisse. 32(2):135-154.
Poulton, J.R.S. (1970). Chemical protection of wine against oxidation. Die Wynboer 466:July:22-23.
Pulley, G.N. and von Loesecke, H.W. (1939). Gases in the commercial handling of citrus juices. Ind. Eng. Chem. 31, 1275-1278.
Rahn, O. and Conn, J.E. (1944). Effect of increase in acidity on antiseptic efficiency. Ind. Eng. Chem. 36:185-187.
Rankine, B.C. (1966). Sulphur dioxide in wines. Food Technol. Australia. 18:134-135, 137, 139, 141.
Rankine, B.C. (1968). The importance of yeasts in determining the composition and quality of wines. Vitis 7, 22-49.
Rankine, B.C. (1995). Making Good Wine. Pan Macmillan Australia Pty Limited. (First published by The Macmillan Co. of Aust., 1989.)
Rankine, B.C. and Pocock, K.F. (1969). Influence of yeast strain on binding of SO2 in wine and on its formation during fermentation. J. Sci. Food and Agric. 20, 104-109.
Rehm, H.-J. and Wittman, H. (1962.) Beitrag zur Kenntnis der antimikrobiellen Wirkung der schwefligen Saure. I. Uebersicht uber einflussnehmende Faktoren auf die antimikrobiellen Wirkung der schwefligen Saure. Z. Lebensm.-Untersuch. -Forsch. 118:413-429.
Rhem, H.J. (1964). The antimicrobial action of sulphurous acid. In Microbial Inhibitors in Food, Ed. Molin, N. Stockholm, Sweden: Almquist, and Wiksells.
Rhem, H.J., Wallnofer, P., and Wittman, H. (1965). Beitrag zur Kenntnis der antimikrobeillen Wirkung der schwefligen Saure. IV. Dissoziation und antimikrobeille Wirkung einiger Sulfonate. Z. Lebens. Forsch. 127:72-85.
Ribéreau-Gayon et al. 1975. Traite d'Oenologie. Dunod, Paris, Tome 2, 267-269.
Ripper, M. (1892). Schweflige Säure in Weinen und deren Bestimmung. J Prakt Chem 46:428-473.
Romano, P. and Suzzi, G. (1992). in Wine Microbiology and Biotechnology (ed. G.H. Fleet). Harwood Academic Publishers, Chur. Switzerland.
Romano, P. and Suzzi, G. (1993). Sulfur dioxide and wine microorganisms. In: Fleet, G.H. (Ed.), Wine Microbiology and Biotechnology, Harwood Academic Publishers, Chur, Switzerland, pp. 373-393.
Sayavedra-Soto, L.A. and Montgomery, M.W. (1986). Inhibition of polyphenoloxidase by sulfite. J. Food Sci. 51:1531-1536.
Scardovi, V. (1951). Studi sulla resistenza all'anidride solforosa. Nota I - Annali di Microbiolognia, 4, 131.
Scardovi, V. (1952). L'ibribazione e le mutazioni dei lieviti. IX Congresso Internazionale di Industrie Agrarie, Roma.
Scardovi, V. (1953). Studi sulla resistenza dei lieviti all'anidride solforsa. Influenza del NaHSO3 su alcune funzioni matboliche del ceppo originario e del ceppo assuefatto. Ann. Microbiol. 5:140-161.
Schaeffer, A. 1987. Cold temperature sulfiting and stabilization of wines in Alsace. Vineyard & Winery Management 39, May/June.
Schanderl, H. (1959). Die Mikrobiologie des Mostes und Weines. 2nd Ed. Stuttgart, Germany: Eugen Ulmer.
Schimz, K.-L. (1980). The effect of sulfite on the yeast Saccharomyces cerevisiae. Arch. Microbiol. 125:28-95.
Schroeter, L.C. (1966). Sulfur dioxide application in foods, beverages, and parmaceuticals. New York, Pergamon Press.
Singleton, V.L. (1987). Oxygen with phenols and related reactions in musts, wines and model systems: Observations and practical implications. Am. J. Enol. Vitic, 38:69-77.
Singleton, V.L., Zaya, J., and Trousdale, E. (1980). White table wine quality and polyphenol composition as affected by must sulfur dioxide content and pmace contact time. Am. J. Enol. and Vitic. 31(1):14-20.
Sudraud, M. (1977). Trate d'Oenologie. Tome IV.
Sudraud, P. and Chauvet, S. (1985). Activité antilevure de l'anhydride sulfureux moléculaire. Conn. Vigne Vin. 19, (1) 31-40.
Suzzi, G., Romano, P., Zambonelli, C. (1985). Saccharomyces strain selection in minimizing SO2 requirement during vinification. Am. J. Enol. and Vitic. 36, 199-202.
Tartar, W.V. and Garretson, H.H. (1941). The thermodynamic ionization constants of sulfurous acid at 25 degrees. J. Am. Chem. Soc. 63:808-816.
Timberlake, C.F. and Bridle, P. (1976). Interactions between anthocyanins, phenolic compounds, and acetaldehyde and their significance in red wines. Am. J. Enol. and Vitic. 27:97-105.
Traverso-Rueda, S. and Singleton, V.L. (1973). Catecholase activity in grape juice and its implications in winemaking. Am. J. Enol. Vitic. 24:103-109.
Tressler, D. K., et al. (1980). Fruit and Vegetable Juice Processing Technology. AVI Publishing Co. 3rd ed.
Usseglio-Tomasset, L. (1989). Chimie oenologique. Tec and Doc Lavoisier, Paris in Ribéreau-Gayon, P., Dubourdieu, D., Donèche, B., Lonvaud, A. (2000). Handbook of Enology, Volume 1, The Microbiology of Wine and Vinifications, John Wiley & Sons Ltd. (English edition), p 183.
Usseglio-Tomasset, L. and Bosia, P.D. (1984.) La prima costante di dissociazione dell'acido solforoso. Vini d'Italia 5, 7-14.
Uzuka, Y. and Nomura, T. (1986). Determination of sulfite resistance in wine yeasts. Proc. 6th Aust. Wine Ind. Tech. Conf., T.H. Lee, Ed., p141-145.
Weger, B. (1956). On desulfiting apple juices. Die Fruchtsaft-Industrie, 5, 212-213.
White, B.B and Ough, C.S. (1973). Oxygen uptake studies on grape juice. Am. J. Enol. Vitic. 24:148-152.
Wildenradt, H.L. and Singleton, V.L. (1974). The production of aldehydes as a results of oxidation of polyphenolic compounds and its relation to wine aging. Am. J. Enol. Vitic. 25:119-126.
Würdig, G. and Schlotter, H.A. (1967). SO2-Bildung in Gärenden Traubenmosten. Z.Lebensm.-Untersuch.-Forsch. 134:7-13.
Yang, H.Y. (1975). Effect of sulfur dioxide on the activity of Schizosaccharomyces pombe. Am. J. Enol. Vitic. 26:1.
Zang, K. and Franzen, K. (1966). Schweflige-Saure-Bildung im Verlauf der Traubenmost-Garung. Deut. Wein-Ztg. 102:128,130.
Zang, K. and Franzen, K. (1967). Bildung von schwefligen Säure bei der Weinbereitung und ihre möglichen Ursachen. Deut. Wein-Ztg. 102:128, 130.
Zoecklein, B., Fugelsang, K.C., Gump, B.H., and Nury, F.S. (1990). Production Wine Analysis. Van Nostrand Reinhold, New York.
Zoecklein, B. (2005). Enology Notes # 101. April 20, 2005.
Zoecklein, B. (2001). Enology Notes #26, August 29, 2001.
Zoecklein, B. (2000). Harvest Memo #4, September 27, 2000.