26.1
Cold Stabilization

26.1.1 Introduction

In winemaking and other areas of food production, stability is the attainment of a state or condition in which the product will have an acceptably low risk of demonstrating undesirable physical or sensory changes over a specified storage time and under a defined set of storage conditions.1 In wine, instabilities that lead to visible changes to the wine are of particular concern to winemakers since they are readily detectable by consumers and lead to unnecessary concerns about product safety. The two most important visible defects that can occur during storage are tartrate instabilities (discussed in this chapter) and protein instabilities (Chapter 26.2).

The chemistries of free tartaric acid (H₂T) and its conjugate bases (bitartrate (HT⁻), tartrate (T²⁻)) were discussed in Chapter 3. The total concentration of tartrate species is usually in the range of 2–6 g/L in wine (0.013–0.040 M). The relative distribution of the tartrate species is pH dependent, but across the range of typical wine pH values, HT⁻ represents the major species (50–70%) and reaches a maximum around pH 3.65 (Chapter 3, see Figure 3.2). Beyond affecting pH and TA, HT⁻ can react with K⁺, the major metal cation in wine (0.003–0.07 M, Chapter 4), to form the poorly soluble potassium bitartrate (KHT, “cream of tartar”). Because the solubility of KHT is lower in ethanolic solutions than in water, and grape juice is nearly saturated in KHT, it is common to see the formation of KHT crystals along tank and barrel walls during the fermentation process; these can be readily separated from the wine by racking. As discussed below, KHT precipitation will decrease TA but may lead to either an increase or decrease in pH due to the amphoteric nature of HT⁻. The low solubility of KHT also indirectly limits the maximum total concentration of tartaric acid species.

The majority of wines exist as metastable, supersaturated KHT solutions due to the presence of crystallization‐inhibiting compounds and other factors. Colder temperatures decrease KHT solubility and accelerate crystal growth, and the potential for visible crystal formation increases upon chilling. Thus, the metastable state could be lost (and KHT crystallization observed) at a consumers’ refrigerator temperature. The resulting KHT crystals are harmless,² but because of their resemblance to broken glass they place a product at risk of consumer rejection. Several routine tests have been developed to predict if a wine is cold‐stable, and most wineries will attempt to cold‐stabilize their wines through one or more methods prior to bottling and commercial sale, particularly for white wines where refrigeration is more common and the resulting crystallization more evident.

Although most studies of KHT metastability focus on the phenomenon during cold storage, KHT precipitation can occur at room temperature due to changes in wine chemistry that are either intentional (e.g., pH adjustments) or spontaneous (e.g., precipitation or degradation of inhibiting compounds). Beyond KHT, other salts of organic acids can precipitate in bottle, particularly calcium tartrate (CaT).

26.1.2 KHT crystal properties and solubility

KHT forms orthorhombic crystals when precipitated from water or model hydroalcoholic solutions. Orthorhombic crystals appear as cubic crystals stretched along two of three axes, and can have several characteristic (idiomorphic) shapes (Figure 26.1.1a), including rectangular prisms and diamonds (hence the term “wine diamonds”).

Figure 26.1.1 Light microscope images of (a) idiomorphic potassium bitartrate crystals from grape juice showing minimal co‐precipitation and potassium bitartrate crystals from red grape juice and wine showing (b) irregular platelets and (c) rounding of crystal edges caused by phenolics and other wine components.

(a) and (b) are from Alongi 2010 [1], reproduced with the permission of Journal of Agriculture and Food Chemistry. (c) is adapted from Anonymous 2015 [2], reproduced with the permission of Australian Wine Research Institute

The solubility of KHT is defined by its solubility product (K _sp), the product of ion concentrations, and is a constant for a particular salt in a given matrix and temperature. In other words, K _sp can be used to calculate the maximum concentration of a salt that can be dissolved in wine or another solvent before saturation is reached. For KHT, the calculation can be expressed as follows:

(26.1.1)

In real solutions, like wine, it is more appropriate to discuss the “activity” or “effective concentration” of a species rather than its concentration, where activity ≰ concentration ([X]). The activity coefficients, γ, are calculated as the ratio of activity to concentration. A modified version of the solubility product can be expressed in terms of activities:

(26.1.2)

For dilute concentrations of ions in pure water, γ = 1. In real solutions, γ is < 1 due to interactions of ions with oppositely charged ions, and lower values of γ are generally observed at higher ionic strength, a concept discussed in more detail earlier in the book (Chapter 3). For example, K⁺ activity is approximately 20–30% lower (i.e., γ = 0.7–0.8) when measured by an ion selective electrode (which measures activity) as compared to [K⁺] measured by atomic absorption spectroscopy (which disrupts weak interactions and measures concentration) [3]. K⁺ may form weak interactions with sulfate and anthocyanins [4], while HT⁻ or T²⁻ may weakly complex with cations in solution, for example, Ca²⁺ to form solubilized CaT [5].

In practice, K _sp is determined by adding excess KHT to a solvent to create a saturated solution, measuring [K⁺] using an appropriate technique (e.g., atomic absorption or emission spectroscopy), and taking the square root to calculate K _sp (since [K⁺] = [HT⁻] upon dissolution of any KHT). Thus, most literature K _sp values of KHT incorporate any effects of the species under investigation (K⁺, HT⁻) on activity, although they will not account for effects of other ions.

The K _sp value for KHT in H₂O at room temperature is 1.07 × 10⁻³, which equates to a solubility of 6.2 g/L. Tables of K _sp values and KHT solubility as a function of temperature and ethanol content can be found in the literature [6], and representative values are shown in Table 26.1.1. Key factors governing KHT solubility are:

• K _sp and solubility decrease with increasing ethanol concentration. The solubility of KHT is usually about 2–3‐fold lower in model wine solutions than in water.
• Colder temperatures decrease K _sp and solubility. Typically, solubility decreases by a factor of 2–3 when a wine is chilled from 20 °C and 0 °C.

Table 26.1.1 KHT solubility in representative matrices, expressed as the solubility product (K_sp) and as solubility in g/L (top), and typical concentration product (CP) values for cold‐stable wines

Matrix	K sp (mol2/L2)	Solubility (g/L)
Water, 20 °C	107 × 10−5	6.2
12% (v/v) ethanol in water, 20 °C	16 × 10−5	2.4
12% (v/v) ethanol in water, 0 °C	2.9 × 10−5	1.0
Matrix	Average CP (mol2/L2)
White wine, stable at 0 °C	4.1 × 10−5
Red wine, stable at 0 °C	7.7 × 10−5

26.1.2.1 Concentration products and KHT metastability

For individual wines, it is possible to calculate the concentration product (CP) in a manner analogous to calculation of K _sp.³

(26.1.3)

These CP values can then be used to evaluate the susceptibility of a wine to precipitating KHT, such that:

• If CP > K _sp, the wine is supersaturated, and further precipitation of KHT is possible, although not guaranteed for reasons described later.
• If CP = K _sp, the wine is saturated, and no additional precipitation or dissolution is possible.
• If CP < K _sp, the wine is not saturated, and further dissolution of KHT would be possible (i.e., if it is added).

As mentioned above, [K⁺] is relatively straightforward to measure directly by spectroscopic methods. The concentration of [HT⁻] is usually not directly measured, but instead approximated by measuring total tartaric acid species, [Tartrate],⁴ and then correcting for the fraction expected to exist as the HT⁻ species based on pH and ethanol determinations for the wine. Tables reporting %HT⁻ as a function of pH and ethanol concentration exist in the literature [6]. From these measurements, CP is then calculated:

(26.1.4)

Surprisingly, the CP values of stabilized commercial wines are comparable to or greater than K _sp values for KHT in analogous ethanol–water model solutions (Table 26.1.1), indicating that most wines are supersaturated in KHT and exist as metastable KHT solutions. The challenge to a winemaker is to ensure that this metastability persists until the point that the consumer opens the bottle – that is, several months at least, and potentially a number of years.

26.1.3 Critical factors for KHT precipitation

The process of formation of a visible KHT precipitate during storage, as summarized in Figure 26.1.2, involves three variables:

• Supersaturation based on ion activities
• Nucleation
• Crystal growth

Figure 26.1.2 Summary of steps necessary for KHT precipitation. Even when a wine is supersaturated with KHT, precipitation may be inhibited by other wine components (A) that decrease K⁺ and HT⁻ activities or (B) that inhibit KHT nucleation and/or crystal growth

26.1.3.1 Ion concentration versus ion activity

The solubility product is properly defined based on the product of ion activities, but, as mentioned earlier, ion activities are rarely measured during wine analyses even when feasible.⁵ In the case of HT⁻, appropriate techniques for measurement of activity in wine rather than concentration have not been described. The fact that concentration is greater than activity is one reason measured CP values can exceed K _sp values.

26.1.3.2 Induction period and nucleation

KHT supersaturation creates the possibility of precipitation, but does not guarantee its occurrence. KHT crystals must first nucleate – that is, form crystals of sufficient size that they can grow further. Initially, a supersaturated solution will form small “embryos” of KHT crystals [7]. Formation of these embryos is usually energetically unfavorable – the small crystals will have a high surface‐to‐volume ratio and the surface KHT molecules will not be as tightly bound as KHT molecules in the interior.⁶ As a result, these embryos will be more likely to dissolve than to grow (Figure 26.1.3). However, a small percentage of embryos may eventually get large enough (i.e., have a sufficiently small surface‐to‐volume ratio) such that continued growth will result in a decrease in free energy. This minimum size is referred to as the critical nucleus radius and the transition is referred to as nucleation. The time prior to nucleation is referred to as the induction period, although commonly the induction time is determined by when crystals become visible and therefore includes both the induction period and subsequent crystal growth. In practical terms:

• The rate of nucleation increases as the exponent of (–1/σ ²), where σ is the degree of supersaturation (calculated as ln(CP/K _sp)). Chilling wine decreases solubility of KHT and results in an increase in the nucleation rate and rate of KHT loss by increasing σ [8].
• In pure solutions, KHT nucleation can occur due to spontaneous growth of embryos in the bulk solution (“homogeneous nucleation”). However, it is much more common in wine and other systems for nucleation events to occur along surfaces (“heterogeneous nucleation”) rather than in the bulk solution. This minimizes the crystal surface area and decreases the likelihood of redissolution of embryos. Thus, KHT crystals are frequently found in imperfections along the walls of tanks or barrels, or on components adhering to these walls, for example, yeast lees⁷ [9].
• Addition of KHT seed crystals (“contact seeding”, i.e., the contact process) is frequently used to expedite crystal growth. Assuming the same mass is used, smaller seed crystals, for example, 20 μm in place of 100 μm crystals, result in faster loss of KHT due to their greater surface area.

Figure 26.1.3 A representation of free energy versus crystal size. Growth of crystals smaller than the critical radius is energetically unfavorable

26.1.3.3 Crystal growth and surface fouling

Critical nucleus radii are typically on the order of 1–100 nm [7]. These are invisible to the human eye and are also several orders of magnitude smaller than typical crystal sizes recovered in wines. Following nucleation, KHT crystals are hypothesized to grow primarily by the separate addition of solvated K⁺ and HT⁻ ions, rather than addition of neutral KHT units to the crystal [10]. The ions are first weakly adsorbed on to the crystal surface before they are eventually incorporated into the lattice, with incorporation of HT⁻ slower than K⁺ due to its greater size and degree of solvation, and slower dehydration. Several factors will affect crystal growth rates:

• The rate of crystal growth in wines and juices is enhanced by agitation [11], either by facilitating transfer of ionized species to the surface or by facilitating dehydration (i.e., removal of solvent from the crystal surface).
• As with nucleation, higher degrees of supersaturation will yield faster crystal growth.
• Crystal growth can be slowed by crystal surface fouling or the adsorption of a wide range of molecules to KHT surfaces that inhibit further growth.

Crystal fouling is a major cause of the metastability and long induction periods of wines as compared to ethanolic solutions. Several species have been implicated for their ability to adsorb to crystal surfaces through ionic, hydrogen‐bonding, and/or charge‐transfer interactions [10, 12], as follows:

• Endogenous phenolics such as anthocyanins, which explains the greater KHT holding capacity of red wines [13].
• Endogenous macromolecules (e.g., proteins, polysaccharides), particularly mannoproteins, which explains the greater KHT holding capacity of sparkling and sur lie (lees) aged wines [14].
• Exogenous components added for the purpose of stabilizing the wine (mannoproteins again, as well as gum arabic, carboxymethylcellulose, and metatartaric acid), as mentioned later in this chapter.

Many of these stabilizing agents are macromolecules (e.g., polysaccharides), which will form colloidal dispersions both in isolation and following binding to KHT; as a result, it is common in the literature to see these stabilizers referred to as “protective colloids”.⁸

In addition to slowing crystal growth, adsorption of fouling compounds will also change KHT crystal morphology due to the preferential binding to one or more of the crystal faces (Figure 26.1.1). For example, the presence of mannoproteins and other wine polysaccharides results in a rounding of crystal faces in white wines [10], while the presence of anthocyanins results in irregular platelets [9]. In fact, the amount of desirable wine components lost due to co‐precipitation can be sizable. For example, KHT crystals from Carignan wines are reported to contain 0.2–0.3 % w/w anthocyanins and 1.9–2.5 % w/w tannin [9], and cold‐stabilization of Concord grape juice can yield KHT crystals with >1% w/w anthocyanins and result in >30% loss of anthocyanins from the juice [1].

The effects of seed crystals and protective colloidal compounds on the rate and extent of KHT precipitation can be seen in Figure 26.1.4.

Figure 26.1.4 The effect of KHT seeding and macromolecule removal (achieved through ultrafiltration, UFT) on KHT precipitation from grape must. A decrease in conductivity indicates that K⁺ and HT⁻ ions have been lost from solution. Addition of 4 g/L of seed crystal results in a large increase in the rate of precipitation and decreases the time necessary to achieve equilibrium. UFT removes macromolecular colloidal compounds that can inhibit crystal growth past a certain size. Based on data from Reference [15]

26.1.4 Testing for KHT stability

Prior to bottling, most commercial wines are assessed for their potential to form precipitates during storage, and particularly during refrigeration [16]. As described above, it is not appropriate to evaluate cold‐stability by comparing CP values to K _sp due to the presence of crystal‐inhibiting compounds in wine. However, the stability of a wine can be predicted by comparing its CP values to empirical observations generated from a large number of comparable wines – effectively an actuarial approach to predicting cold stability. These maximum CP values will vary with wine type, for example, red wines can have higher CP values due to the presence of phenolic species (Table 26.1.1). The maximum CP values recommended for stability vary with wine style, region, vintage, and variety, but as representative data, CP values for California wines were recommended to be < 16.5 × 10^–5 for table whites and < 30 × 10^–5 for table reds [17]. For lesser known wines or locations, specific CP limits may be unavailable.

An alternate approach to test for tartrate stability is to subject a wine to forcing conditions – that is, expose the wine to conditions that will accelerate the rate of KHT crystal growth. A widely used test involves agitating a wine sample at 0 °C in the presence of KHT seed crystals and measuring changes in conductivity at 5 min intervals for up to 30 min. A conductivity change of > 5% is reportedly indicative of likely KHT instability [17]. Several other variants exist, including visually inspecting for crystal formation rather than conductivity, omitting seeding, using longer or shorter hold times, or using cooler temperatures (Table 26.1.2).

Table 26.1.2 Approaches to testing for KHT cold stability

General approach	Examples or variations [16–20]	Notes
Compare wine CP value to recommended CP limits	Published tables for certain regions. Commercial labs may use proprietary values	Empirically derived predictions. Data not available for all regions or wine styles. Requires measuring [K+], [Tartrate], pH, and ethanol
Chill wine, check for crystal formation (“cold‐hold test”)	Several variants exist: Temperature: typically –4 to 0 °C Time: days to weeks Other variables: agitation	Simple, no specialized equipment required. Common in wineries. May be excessively stringent
Chill wine, monitor for change in conductivity	Several variants exist: Temperature: typically, –4 to 0 °C Time: minutes to hours Other variables: often performed with seeding (“mini‐contact”) and agitation	Faster and more sensitive than visual inspection, but requires conductivity meter
Determine minimum temperature necessary for KHT dissolution (“saturation temperature,” T sat)	Typically, a control sample and sample with added KHT are heated at a controlled rate, and T sat is determined by when conductivity of the treated sample exceeds the control Variants: Heating rate (°C/min) T sat limit: typically > 12.5 °C (white) and 22 °C (red) indicative of instability	Less dependent on crystal size and other parameters than mini‐contact tests. Requires conductivity meter
Freeze wine, thaw, check for crystal formation	Conditions poorly defined. Overnight storage is common. Freezer temperatures can vary from –10 °C to –27 °C resulting in variation [20]	Common in wineries. Due to change of state, potential for false positives and negatives in predicting behavior under refrigeration conditions

As with other stability tests (e.g., protein stability, Chapter 26.2), the goal of these forcing tests is to predict long‐term behavior of the wine (months to years) in an accelerated period (minutes to weeks). These run the risk of errors:

• False positives, especially if the testing conditions are more extreme than would be encountered post‐bottling (e.g., seeding, freezing). This may lead to unnecessary time, energy, and material resources committed to stabilizing a wine that is at low risk for KHT precipitation.
• False negatives, either because of differences in physical state between testing and real conditions, or because other changes may occur to the wine during long‐term room‐temperature storage (e.g., loss of inhibitory compounds).

In summary, cold‐stability is not an absolute concept; rather, cold‐stability tests approximate the risk of a given wine to form visible KHT precipitates over a particular storage regimen, and the winemaker must balance this risk against the costs of preventative treatments.

26.1.5 Treatments for preventing KHT precipitation

If a wine is determined to be at sufficient risk for KHT instability before consumption, there are several approaches a winemaker can use to decrease the likelihood of this occurring, as summarized in Table 26.1.3, and reviewed in more detail elsewhere [16, 21–23].

Table 26.1.3 Strategies for achieving cold stability in wineries

General approach	Examples or variations	Notes
Decrease ion concentrations	Cold storage to induce KHT precipitation: Often with seeding and agitation Batch or continuous Nanofiltration to preconcentrate K+ and HT− and favor KHT precipitation	Batch process is simple, has the lowest wine losses of any technique, and uses readily available winery equipment. Energy intensive, although possibility for energy recovery. Ineffective for CaT instabilities. Equipment for continuous processes or nanofiltration can be expensive
	Ion exchange resins: Polymeric resins capable of exchanging like‐charged species Most commonly as cation exchange (exchange H+ or Na+ for K+, Ca2+) Less common: anion exchange (exchange citrate or OH− for tartrate ions)	Resins must be regenerated with strong acid (or acid and NaCl) for re‐use. Potential for non‐selective losses of flavor and color compounds. May be other effects on sensory properties due to pH and TA changes
	Electrodialysis High‐pressure flow cell with membranes selectively permeable to small charged species	Expensive equipment. Minimal effect on red wine color compared to cold stabilization. Slight changes to [EtOH], TA and pH
Inhibit nucleation or crystal growth to maintain metastability	Polysaccharides (carboxymethylcelluose, gum arabic, mannoproteins) Peptides (polyaspartate) Metatartaric acid	Treatments can be relatively expensive and dosage rate can be challenging to determine. Potential for changes to filterability, especially with polysaccharides Metatartaric will hydrolyze to tartaric acid during wine storage

26.1.5.1 Cold storage

The most straightforward means to decrease the chance of post‐bottling cold instability is to subject the wine to forcing conditions – that is, store the wine at close to its freezing point to induce loss of KHT [18]. Subject to the prices of resources like energy and the price of the wine, cold storage may also be the most economical [22]. In addition to colder temperatures, the rate of precipitation can be enhanced by agitation and, in particular, the introduction of KHT seed crystals [24], as discussed above. Precipitated KHT can be pulverized and re‐used to treat subsequent batches of wine.

Cold treatments generally yield wines that still have CP > K _sp, and thus continue to be metastable and at potential risk for KHT precipitate formation during subsequent storage. Further decreases in CP can be achieved by physiochemical treatments to remove K⁺ and/or HT⁻.

26.1.5.2 Ion‐exchange resins

Polymeric resin beads (e.g., styrene–divinylbenzene copolymer) containing functional groups capable of exchanging like‐charged ions can be used for decreasing ion species [21]. In winemaking operations, most resins are cation‐exchange – that is, they contain negatively charged residues (sulfonic acid groups, –SO₃H, or carboxylic acid groups, –COOH, respectively giving strong and weak exchange capacities). A cation exchange resin can be prepared by washing it with strong acid prior to use, and wines will then exchange K⁺ for H⁺ in the cartridge. Alternatively, if no change in pH or TA is desired, the cartridge can be rinsed with a solution of a metal salt (e.g., NaCl or MgCl₂) prior to use to allow exchange of K⁺ for Na⁺ or Mg²⁺, respectively [25].⁹ The relative affinity of metals for resins varies between strong and weak exchanges.

• Weak exchangers generally show a preference for higher charge states (Ca²⁺ > K⁺). Within a charge state, smaller ions (higher on the periodic table) are preferred, that is, greater charge density.
• Strong exchangers generally show a preference for lower charge states (K⁺ > Ca²⁺). Within a periodic table column, the higher number element will be preferred since they will have a smaller hydrated diameter.

An example specification sheet for a commercial cation‐exchange resin is shown in Table 26.1.4. Frequently, these resins are sold to winemakers pre‐packaged in cartridges. A critical parameter for evaluating the cost‐effectiveness of a resin is its total exchange capacity expressed as charge equivalents per L of resin. The resin shown in Table 26.1.4 has a capacity of 1.7 eq/L; based on charge and atomic (or molecular) weight this equates to either 66 g/L of monovalent K⁺ cation (i.e., 1.7 × 39) or 34 g/L of the divalent Ca²⁺ cation (i.e., 1.7 × 40/2).

Table 26.1.4 Example data sheet for a commercial cation exchange resin (Dow Amberlite®)

Matrix	Crosslinked polystyrene
Functional groups	Sulfonic acid
Physical form	Light grey beads
Ionic form as shipped	H+
Total exchange capacity	=1.7 eq/L (H+ form)
Harmonic mean size	0.600–0.800 mm

In practice, to avoid the need for precise calculations, it is common to treat only a portion of juice or wine with an excess of cation exchange resin. Under these conditions, >75% of metals (K⁺, Ca²⁺, etc.) will be removed, resulting in an increase in TA in the treated juice corresponding to the initial concentration of metal cation equivalents and a decrease in pH to near 2 [26]. This treated portion can then be back‐blended. Because cation‐exchange resins contain both hydrophobic and negatively charged regions, they can cause non‐selective losses of other wine components, for example, the flavylium form of anthocyanin pigments or flavor compounds.

26.1.5.3 Electrodialysis

Another approach to decreasing ion concentrations is electrodialysis (ED), a technique widely used in food processing for removal or concentration of ionic compounds in foods or waste streams [27, 28].¹⁰ An ED device consists of a central cell containing the product to be electrodialyzed (e.g., wine) separated from anode and cathode compartments, in which water flows, by anion‐ and cation‐selective membranes. When an electric field is applied, ions in wine will migrate towards the appropriate electrode [26]:

• Bitartrate and tartrate will migrate through the anion‐permeable membrane towards the anode. Sulfate (SO₄ ²⁻) will be lost preferentially due to its small size and double charge. Malate/bimalate and lactate will also migrate, but to a lesser extent due to their higher pK_a values.
• Cations, primarily K⁺ but also Mg²⁺, Ca²⁺, and others, will migrate towards the cathode through a cation‐permeable membrane.

Similar to ion exchange, ED processing is typically done in a batch manner with a portion of wine being treated until the desired conductivity is reached, and then discharged and replaced with another batch. The net effect of electrodialysis is superficially similar to cold‐stabilization – there will be a decrease in KHT. However, ED (like ion exchange) will remove Ca²⁺ and other metals as well [29], and thus is potentially more effective at preventing CaT instability. In addition, the removal of sulfate may affect the stability of other wine components, such as proteins (Chapter 26.2) and the elimination of some T²⁻ in addition to HT⁻ results in a slight pH decrease, ~0.2 pH units, in contrast to cold stabilization where either a decrease or increase can occur (Chapter 3).

Finally, a variation on conventional electrodialysis is bipolar electrodialysis (BPED), which incorporates bipolar membranes capable of allowing both anionic and cationic exchange [28]. BPED systems can achieve H⁺ and OH⁻ exchange, which means that they can be designed to selectively remove either free acids (to prevent a pH decrease and minimize TA) or metal hydroxide salts (to increase TA and decrease pH, similar to cation exchange resins). The approach is widely used in the fruit juice industry and has also been applied to wine [30].

26.1.5.4 Additives for preventing nucleation and crystal growth

Certain additives can be used to prevent KHT crystal formation [14] (Table 26.1.3), although their legality varies among countries (Chapter 27). Generally, these work by inhibiting nucleation or fouling crystal faces to prevent formation of visible crystals, as described earlier. Many of these additives are large macromolecules, especially polysaccharides.

26.1.6 CaT and related precipitates

While KHT precipitation is the most common (and best studied) cause of in‐bottle crystalline deposits, others are occasionally observed. For example, calcium oxalate precipitation has been reported to result following addition of tartaric acid contaminated with oxalic acid, and calcium mucate has been reported in wines with botrytis infections. High concentrations of ferric ions can lead to ferric phosphate precipitation (iron casse) as described elsewhere in this book (Chapters 4 and 26.2).

A more common crystalline instability is CaT, most commonly observed when calcium salts are used for deacidification (Chapter 3) or when wines are stored in concrete tanks (Chapter 4). The solubility of CaT is lower than KHT, 0.1 versus 2.4 g/L at 20 °C in 12% v/v ethanol. Similar to KHT, CaT often exists in supersaturation in wines, with CP > K _sp by a factor of 2–5 [31]. Several components in wine are capable of inhibiting CaT nucleation or crystal growth, including malic acid and uronic acids (i.e., pectin) [32]. However, the occurrence of CaT instability is usually lower because [T²⁻] (i.e., divalent tartrate required to form CaT) is generally lower than [HT⁻] at wine pH, and because [K⁺] is usually 10‐fold higher than [Ca²⁺] in grapes (Chapter 4). Additionally, Ca²⁺ activity in wine is decreased in the presence of T²⁻, and a portion of CaT will be molecularly dispersed in solution without precipitating [5]. In contrast to KHT, the rate and amount of CaT precipitation from wine is virtually unaffected by temperature [33], and cold‐storage is thus not an appropriate strategy for CaT stabilization or for testing CaT stability of wine. Seeding and agitation greatly increase the rate of CaT precipitation in wine [34], indicating that the nucleation rate is likely a limiting step, but seeding is not often used in wineries due to the unavailability of pure seed crystals. Ion‐exchange resins or electrodialysis are appropriate ways to lower [Ca²⁺] and [T²⁻], but in practice the best way to avoid CaT precipitation is to avoid the use of calcium salts during wine production.

References

1. 1. Alongi, K.S., Padilla‐Zakour, O.I., Sacks, G.L. (2010) Effects of concentration prior to cold‐stabilization on anthocyanin stability in Concord grape juice. Journal of Agricultural and Food Chemistry , 58 (21), 11325–11332.
2. 2. Anonymous (2015) Hazes and deposits picture gallery. Australian Wine Research Institute. Available from: https://www.awri.com.au/industry_support/winemaking_resources/wine_instabilities/hazes_and_deposits/picture_gallery/.
3. 3. Gerbaud, V., Gabas, N., Blouin, J., Laguerie, C. (1996) Nucleation studies of potassium hydrogen tartrate in model solutions and wines. Journal of Crystal Growth , 166 (1–4), 172–178.
4. 4. Bertrand, G.L., Carroll, W.R., Foltyn, E.M. (1978) Tartrate stability of wines. I. Potassium complexes with pigments, sulfate, and tartrate ions. American Journal of Enology and Viticulture , 29 (1), 25–29.
5. 5. McKinnon, A.J., Scollary, G.R., Solomon, D.H., Williams, P.J. (1994) The mechanism of precipitation of calcium L(+)‐tartrate in a model wine solution. Colloids and Surfaces A – Physicochemical and Engineering Aspects , 82 (3), 225–235.
6. 6. Berg, H.W. and Keefer, R.M. (1958) Analytical determination of tartrate stability in wine I. Potassium bitartrate. American Journal of Enology and Viticulture , 9 (4), 180–193.
7. 7. De Yoreo, J.J. and Vekilov, P.G. (2003) Principles of crystal nucleation and growth. Reviews in Mineralogy and Geochemistry, 54 (1), 57–93.
8. 8. Dunsford, P. and Boulton, R. (1981) The kinetics of potassium bitartrate crystallization from table wines II. Effect of temperature and cultivar. American Journal of Enology and Viticulture , 32 (2), 106–110.
9. 9. Vernhet, A., Dupre, K., Boulange‐Petermann, L., et al. (1999) Composition of tartrate precipitates deposited on stainless steel tanks during the cold stabilization of wines. Part II. Red wines. American Journal of Enology and Viticulture , 50 (4), 398–403.
10. 10. Rodriguez‐Clemente, R. and Correa‐Gorospe, I. (1988) Structural, morphological, and kinetic aspects of potassium hydrogen tartrate precipitation from wines and ethanolic solutions. American Journal of Enology and Viticulture , 39 (2), 169–179.
11. 11. Dunsford, P. and Boulton, R. (1981) The kinetics of potassium bitartrate crystallization from table wines. I. Effect of particle size, particle surface area and agitation. American Journal of Enology and Viticulture , 32 (2), 100–105.
12. 12. Celotti, E., Bornia, L., Zoccolan, E. (1999) Evaluation of the electrical properties of some products used in the tartaric stabilization of wines. American Journal of Enology and Viticulture , 50 (3), 343–350.
13. 13. Balakian, S. and Berg, H. (1968) The role of polyphenols in the behavior of potassium bitartrate in red wines. American Journal of Enology and Viticulture , 19 (2), 91–100.
14. 14. Marchal, R. and Jeandet, P. (2009) Use of enological additives for colloid and tartrate salt stabilization in white wines and for improvement of sparkling wine foaming properties, in Wine chemistry and biochemistry (eds Moreno‐Arribas, M.V. and Polo, M.C.), Springer, pp. 127–158.
15. 15. Bott, E. (1986) Centrifugal separation of tartrate from wines stabilized by the contact process. Wine Industry Journal, (August), 35–38.
16. 16. Howe, P. (2013) Potassium bitartrate/calcium tartrate: cold stability of wines, Part 2. Practical Winery and Vineyard, (Winter), 34–39.
17. 17. Zoecklein, B.W., Fugelsang, K.C., Gump, B.H., Nury, F.S. (1999) Wine analysis and production , Kluwer Academic/Plenum Publishers, New York.
18. 18. Iland, P., Bruer, N., Ewart, A., Markides, A., Sitters, J. (2004) Monitoring the winemaking process from grapes to wine techniques and concepts. Patrick Iland Wine Promotions Pty, Ltd. Adelaide, SA, Australia.
19. 19. dos Santos, P.C., Gonçalves, F., De Pinho, M.N. (2002) Optimisation of the method for determination of the temperature of saturation in wines. Analytica Chimica Acta , 458 (1), 257–261.
20. 20. Leske, P.A., Bruer, N.G.C., Coulter, A.D. (eds) (1995) Potassium tartrate – how stable is stable? 9th Australian Wine Industry Technical Conference, 1995, Adelaide, SA, Australia.
21. 21. Lasanta, C. and Gómez, J. (2012) Tartrate stabilization of wines. Trends in Food Science and Technology , 28 (1), 52–59.
22. 22. Low, L.L., O’Neill, B., Ford, C., et al. (2008) Economic evaluation of alternative technologies for tartrate stabilisation of wines. International Journal of Food Science and Technology , 43 (7), 1202–1216.
23. 23. Bosso, A., Panero, L., Petrozziello, M., et al. (2015) Use of polyaspartate as inhibitor of tartaric precipitations in wines. Food Chemistry , 185, 1–6.
24. 24. Boulton, R.B., Singleton, V.L., Bisson, L.F., Kunkee, R.E. (1999) Principles and practices of winemaking , Kluwer Academic/Plenum Publishers, New York.
25. 25. Mira, H., Leite, P., Ricardo da Silva, M., Curvelo‐Garcia, A. (2006) Use of ion exchange resins for tartrate wine stabilization. Journal International Des Sciences de la Vigne et du Vin , 40 (4), 223–246.
26. 26. Escudier, J.L., Cauchy, B., Lutin, F., Moutounet, M. (2012) Acidification and tartaric stabilization. Technological comparison of ion exchange resins by extraction and ion membrane. Progres Agricole et Viticole , 129 (13/14), 324–332.
27. 27. Hestekin, J., Ho, T., Potts, T. (2010) Electrodialysis in the food industry, in Membrane technology , Wiley‐VCH Verlag GmbH & Co. KGaA, pp. 75–104.
28. 28. Mondor, M., Ippersiel, D., Lamarche, F. (2012) Electrodialysis in food processing, in Green technologies in food production and processing (eds Boye, J.I. and Arcand, Y.), Springer US, pp. 295–326.
29. 29. Gómez Benı&c.acute;tez, J., Palacios Macı&c.acute;as, V.M., Szekely Gorostiaga, P., et al. (2003) Comparison of electrodialysis and cold treatment on an industrial scale for tartrate stabilization of sherry wines. Journal of Food Engineering , 58 (4), 373–378.
30. 30. Rozoy, E., Bazinet, L., Gagne, F., et al. (2013) Development of a deacidification of wine by a bipolar membrane electrodialysis: a feasibility study of the laboratory scale. Bulletin de lʼOIV , 86 (986/987/988), 187–208.
31. 31. Berg, H.W. and Keefer, R.M. (1959) Analytical determination of tartrate stability in wine. II. Calcium tartrate. American Journal of Enology and Viticulture, 10 (3), 105–109.
32. 32. McKinnon, A.J., Williams, P.J., Scollary, G.R. (1996) Influence of uronic acids on the spontaneous precipitation of calcium L‐(+)‐tartrate in a model wine solution. Journal of Agricultural and Food Chemistry , 44 (6), 1382–1386.
33. 33. Clark, J.P., Fugelsang, K.C., Gump, B.H. (1988) Factors affecting induced calcium tartrate precipitation from wine. American Journal of Enology and Viticulture , 39 (2), 155–161.
34. 34. Abguéguen, O. and Boulton, R.B. (1993) The crystallization kinetics of calcium tartrate from model solutions and wines. American Journal of Enology and Viticulture , 44 (1), 65–75.

Notes

1 Note that stability does not imply that a product will have zero risk of undesirable changes. Similar to safety, product stability should be thought of as a probability of failure, not the impossibility of failure.

2 Although the dose eventually does make the poison. “This cream of tartar case is only another illustration of how necessary it is that some law should be passed for not allowing incompetent people to sell drugs…[it has been] only a few weeks since a grocer sold cream of tartar in mistake for arrowroot thus causing the death of a child,” (Chemist and Druggist, February 15, 1890, Vol. 36, General and Provincial News: Incompetent drug sellers, p. 211).

3 Concentration products (CPs) are an example of reaction quotients (Q), which can be compared against equilibrium (K) values to predict the direction a reaction will proceed in. General chemistry textbooks typically use Q rather than CP when discussing saturation and other equilibrium phenomena, but the CP terminology persists in usage in many wine texts.

4 Tartrate can be measured in several ways, including by HPLC, and by visible spectroscopy using a metavanadate reagent.

5 The one routine exception is ion selective electrodes such as a pH probe, because pH is technically a measurement of proton activity rather than proton concentration.

6 The difference in free energy between molecules at the surface and interior is referred to as “interfacial free energy” or “surface free energy” in chemistry texts.

7 A more rapid and enjoyable example of heterogeneous nucleation can be seen with the addition of a Mentos candy to a bottle of soda, the results of which can be found through an online search. The high surface area of the candy provides a large number of nucleation sites for the supersaturated CO₂ gas.

8 As a reminder, a colloidal solution is a homogeneous solution of large macromolecules or small particles that does not settle out to form two separate phases.

9 Cation-exchange resins are typically found in homes as part of water-softening units. In this case the primary metals targeted are usually the divalent metals Ca²⁺ and Mg²⁺, which can interfere with soap lathering and generate insoluble scales. These cations are exchanged for Na⁺ on the resin.

10 Other examples of applications in the food industry include demineralization of whey and concentration of plant proteins.