Читать книгу Handbook of Enology, Volume 2 - Pascal Ribéreau-Gayon - Страница 34

The temperature at which a wine becomes capable of dissolving bitartrate is a useful indication of its state of supersaturation. However, in practice, enologists prefer to know the temperature below which there is a risk of tartrate instability. Maujean et al. (1985, 1986) tried to determine the relationship between saturation temperature and stability temperature.

The equations for the solubility (A) and supersolubility (B) curves (Section 1.5.1, Figure 1.11) were established for this purpose by measuring electrical conductivity. They follow an exponential law of the following type: C = ae^bt, where C is the conductivity, t is the temperature, and a and b are constants.

The experiment to obtain the exponential supersolubility curve (B) consists of completely dissolving added cream of tartar in a wine at 35°C and then recording the conductivity as the temperature dropped. This produces an array of straight‐line segments (Figure 1.11) whose intersections with the exponential solubility curve (A) correspond to the saturation temperatures of a wine in which an added quantity i of KHT has been dissolved. The left‐hand ends of these straight‐line segments correspond to the spontaneous crystallization temperatures . For example, if 3 g/l of KHT is dissolved in wine, the straight line representing its linear decrease in conductivity stops at a temperature of 18°C, i.e. the temperature where spontaneous crystallization occurs .

Of course, if only 1.1 g/l of KHT is dissolved in the same wine, crystallization occurs at a lower temperature, as the wine is less supersaturated . It is therefore possible to obtain a set of spontaneous crystallization temperatures based on the addition of various quantities i of KHT (Figure 1.11).

The range covering this set of spontaneous crystallization temperatures defines the exponential supersolubility curve (B). The exponential solubility and supersolubility curves, representing the boundaries of the domain of supersaturation (DS), are parallel. This property, first observed in Champagne base wines, is used to deduce the spontaneous crystallization temperature of the initial wine.

Indeed, when the intersections of the straight conductivity lines with the two exponentials (A) and (B) are projected on the temperature axis, we obtain temperatures and , respectively. The difference, , defines the width of the DS of the wine in which i added KHT has been dissolved, expressed in degrees Celsius. The width of the domain of supersaturation is independent of the addition value i, as exponential curves (A) and (B) are roughly parallel. Thus, in the example described (Figure 1.11), the width of the domain of supersaturation is close to 21°C, whether 1.1 g/l or 1.8 g/l of KHT is added. If 21°C is subtracted from the true saturation temperature of the wine , i.e. no added KHT (i = 0), it may be deduced that spontaneous crystallization is likely to occur in this wine at temperature .

The experimental method for finding the width of the domain of supersaturation has just been described, and the relationship between the saturation temperature and the temperature below which there is a risk of crystallization has been deduced. The width of the domain of supersaturation, corresponding to the delay in crystallization, must be linked, at least partially, to the phenomenon of supercooling (the effect of alcohol), as well as the presence of macromolecules in the wine that inhibit the growth of the nuclei. These macromolecules include carbohydrate, protein, and phenol colloids. It seems interesting, from a theoretical standpoint, to define the contribution of these protective colloids to the width of the domain of supersaturation. It also has a practical significance and should be taken into account in preparing wines for tartrate stabilization. For this purpose, aliquots of the same white wine at 11% vol. alcohol were subjected to various treatments and fining (Table 1.16). At the same time, a model dilute alcohol solution was prepared: 11% vol. buffered at pH 3, containing 4 g/l of KHT, with a saturation temperature of 22.35°C. The spontaneous crystallization temperature of the same solution was also determined after 1.4 g/l of KHT had been dissolved in it, . It was thus possible to find the width of the domain of supersaturation, i.e. 15°C.

The spontaneous crystallization temperature of each sample of treated wine (Table 1.16) was also determined using the same procedure. Examination of the results shows that a wine filtered on a 10³ Da Millipore membrane, i.e. a wine from which all the colloids have been removed, has the lowest value for the domain of supersaturation , closest to that of the model dilute alcohol solution. Therefore, the difference between the results for this sample and the higher values of the domains of supersaturation of fined samples defines the effect of the protective colloids. It is interesting to note that the sample treated with metatartaric acid had the widest supersaturation field and cold stabilization was completely ineffective in this case. This clearly demonstrates the inhibiting effect this polymer has on crystallization and, therefore, its stabilizing effect on wine (Section 1.7.6). Stabilization by this method, however, is not permanent.

On the basis of these results evaluating the protective effects of colloids and saturation temperatures before and after cold stabilization, it is possible to determine the most efficient way to prepare a white wine for bitartrate stabilization. It would appear that tannin–gelatin fining should not be used on white wines, while bentonite treatment is the most advisable. The effect of tannin–gelatin fining bears out the findings of Lubbers et al. (1993), highlighting the inhibiting effect of yeast cell wall mannoproteins on tartrate precipitation.

There are quite tangible differences in the performance of slow stabilization when wines have no protective colloids (wine filtered on a membrane retaining any molecule with a molecular weight above 1,000 Da). These effects ought to be even more spectacular in the case of rapid stabilization technologies. Indeed, the results presented in Figure 1.16 show the impact of prior preparation on the effectiveness of the contact process.

It was observed that the crystallization rate during the first hour of contact, measured by the slope of the lines representing the drop in conductivity of the wine in microsiemens per centimeter per unit time, was highest for the wine sample filtered on a 10³ Da membrane, i.e. a wine containing no protective colloid macromolecules. In contrast, the addition of metatartaric acid (7 g/hl) completely inhibited the crystallization of potassium bitartrate, even after four hours. In production, bentonite and activated charcoal are the best additives for preparing wine for tartrate stabilization using the contact process.