Abstract
The adsorption of sulfur dioxide in the gas phase on crude cork is reported for the first time. Close contact between SO2 and cork occurs either by the liquid phase or by the vapor phase from the headspace, making it crucial to understand the thermodynamic interactions occurring between cork and SO2. Adsorption of gaseous SO2 was studied by thermogravimetry and calorimetry at 298 K and for pressures ranging from 10−5 to 40 hPa on cork sample previously outgassed under vacuum. Amounts of SO2 adsorbed on dry cork are rather low and correspond to negligible quantities when extrapolated to an entire cork stopper. The presence of a hysteresis loop on the isotherm and the high adsorption heats measured at low loading (~100 kJ·mol−1) indicates that a reactive adsorption occurs between SO2 and cork. Nevertheless, the chemisorbed amount is very low and the main adsorption mechanism for SO2 on cork corresponds to a physisorption process. Adsorption of SO2 on partially hydrated cork was also studied. When the water content in cork is of 5% of the weight, the amount of SO2 adsorbed is divided by 3. Thus, water does not enhance the adsorption rate for SO2, but decreases the SO2 adsorption activity onto cork, probably because of competitive adsorption mechanisms.
Sulfur dioxide (SO2) is an anthropogenic global atmospheric pollutant often produced by various industrial processes. It is mainly generated by the combustion of fossil fuels and by incineration of solid waste, which contain sulfur compounds. As a precursor to sulfuric acid (H2SO4), formed by atmospheric oxidation of SO2 in the presence of water, it can then lead to acid rain, as sulfuric acid is extremely soluble in water. Sulfur dioxide is also widely used as antioxidant in foods, including wine. In this empiric use, reported for several centuries, it generally serves for both antiseptic properties and antioxidant effects to protect musts and wines at pressing and bottling, in concentrations from 10 to 500 μg/L.
The study of SO2 adsorption by solid adsorbents was first reported on activated carbons (Shiels 1929), which were proved to be excellent absorbents for SO2 (Arunov et al. 1977, Zhang et al. 2007). This adsorption process has also been studied for zeolites, which exhibit good potentiality as adsorbents for SO2 because of high microporosity (Garcia-Martinez et al. 2002, Kopac 1999). Some studies also report that adsorption of SO2 by activated carbons is mainly governed by the microporosity of the adsorbent (Bagreev et al. 2002, DeBarr et al. 1997). The adsorption phenomenon of SO2 on these adsorbents might have two origins: either physisorption by weak bonds (Arunov et al. 1977), such as hydrogen bonds or ion-dipole bonds (interaction SO2-Na+) in zeolites (Garcia-Martinez et al. 2002), or/and chemisorption involving high-energy bonds (Arunov et al. 1977, Furmaniak et al. 2006, Moreno-Castilla et al. 1993). Therefore the chemical nature and the structure of the surface layer are both determining factors for the adsorption process of SO2, through the adsorption ability of the adsorbent and the accessibility of the adsorbate for the adsorption sites of the adsorbent. In particular, certain surface oxygen groups, such as pyrone structures, may play an important role by increasing the surface basicity of the adsorbents (Kisamori et al. 1994, Moreno-Castilla et al. 1993). Oxidative adsorption in the presence of oxygen and water can also lead to the formation of new adsorbed chemical species derived from SO2, such as SO3 and H2SO4 (Arunov et al. 1977, Bagreev et al. 2002, Furmaniak et al. 2006, Kisamori et al. 1994, Zhang et al. 2007).
Cork, from Quercus suber L., displays a good impermeability to liquids and gases and is traditionally used as stopper for closing wine bottles, in which SO2 is present (Karbowiak et al. in press). The chemical composition of cork as described in literature remains relatively variable and is different both within and between trees (Pereira 1988). Cork is mainly comprised of suberin (33–50% w/w), lignin (13–29%), and polysaccharides (cellulose and hemicellulose, 6–25%). Lignin is thought to be the main constituent of the thin internal primary cell wall, which is surrounded by alternating suberin and wax lamella in the thick secondary wall, which is in turn contained by the thin tertiary wall composed of polysaccharides (Silva et al. 2005). It also contains smaller percentages of waxes (2–8%), tannins (6–7%), extractable (8–24%), ash (2–3%), and other compounds (6–7%) (Silva et al. 2005). The chemical structure of suberin and lignin in cork has not yet been fully elucidated. Suberin is thought to be a macromolecular network of aliphatic polyesters, with various long-chain fatty acids and phenolic moieties, these two domains being spatially distinct (Cordeiro et al. 1998). Because of its specific physical and mechanical properties (Gibson et al. 1981), cork biomaterial is useful in other applications, including in the construction industry as floor and wall covering or as insulation corkboard for thermal insulation, acoustical absorption, and vibration insulation (Silva et al. 2005). Cork also has adsorption properties, and cork powder waste can be used as a biosorbent of pollutants from wastewater as it can be easily incinerated afterward. Moreover, it can be physically or chemically activated in order to obtain activated carbons with molecular sieve properties (Carvalho et al. 2006, Mourao et al. 2006). Cork powder waste has been shown to adsorb various metallic cations from the liquid phase, such as chromium (Cr3+) (Machado et al. 2002), copper (Cu2+) and nickel (Ni2+) (Chubar et al. 2003, Villaescusa et al. 2002), zinc (Zn2+) (Chubar et al. 2003), biphentrin (a pyrethroid) (Domingues et al. 2005), and phenolic compounds (Mourao et al. 2006). Cork can also adsorb various volatile organic compounds (Carvalho et al. 2006). 2,4,6-Trichloroanisole, a wine contaminant known for a cork taste, can easily be absorbed by cork stopper via the vapor phase, mainly confined in the outer 2 mm of the cork cylinder and also with some migration in the inside after 24 hr of exposure to this contaminant (Barker et al. 2001).
This natural organic material therefore displays rather good adsorption capacities for various adsorbates. The adsorption capacity of cork for SO2 has not been studied, however, and it may have enological use both as a stopper and potential application as a biosorbent. The present study therefore focuses on the adsorption of gaseous SO2 by cork at the equilibrium, with particular attention on the effect of the hydration degree of cork on the adsorption properties of gaseous SO2.
Materials and Methods
Material.
Raw cork planks, from Quercus suber L. in the Mora (Portugal) production area, were supplied by the society Bouchons Trescases S.A. (Boulou, France). Planks were not washed or surface treated (with paraffin or silicone). Uniform cork pieces in plate geometry, 35 mm long, 10 mm wide, and 1 mm thick for gravimetric study of SO2 in gas phase, were manually cut from the planks. In this geometry the axial plane has the largest contact area with wine. Cork powder with a ≤500-μm particle size was also used. Sulfur dioxide was used in the gas state (purity 99.9%, molar mass 64.1 g·mol−1; Air Liquide, Paris, France).
Scanning electronic microscopy.
Scanning electronic microscopy (SEM) observations of the cellular cork structure were performed on a JEOL-JSM 6400 F (JEOL, Paris, France), with an acceleration voltage of 15 kV and a preliminary carbon metallization of cork plate samples.
Thermogravimetry.
Adsorption of gaseous SO2 on natural cork was investigated by thermogravimetry with a homemade McBain balance, under controlled conditions of temperature (298 K) and H2O or SO2 gas pressure (controlled with a MKS Baratron absolute pressure transducer; MKS, Le Bourget, France). In this closed glass system, the cork is hung on a quartz helicoidal spring, whose elongation indicates a sample mass variation as a function of the gas pressure at equilibrium. The experimental accuracy is ±0.01 mg for the mass of adsorbate, ±0.5 K for the temperature, and ±1.0 Pa for the pressure. The adsorption-desorption isotherm is measured step-by-step using a static method. Once a plateau of mass is recorded, the next equilibrium is reached by changing the pressure of the gaseous adsorbate. Two different geometries of the sample were studied: cork plate and cork powder. The mass of the cork sample was ~40 mg. The cork sample was outgassed under vacuum (10−5 hPa) at 298 K for 24 hr before each experiment.
Adsorption of gaseous SO2 was studied for pressures ranging from 10−5 to 40 hPa and at 298 K. This corresponds to the relative pressure range 0 ≤ p/ps ≤ 0.0l, the saturated vapor pressure of SO2 (ps) being equal to 4000 hPa at 298 K. Studied pressure range covers the partial pressures of SO2 found in the headspace of a bottle of wine.
Prehydration of the cork sample before SO2 adsorption was also performed by submitting the sample at 298 K to a saturated vapor pressure of water equal to 23 hPa. Under this pressure, the water content at equilibrium was 5% of the weight of the cork. The adsorption of the gaseous sulfur compound was then carried out by successive increments of the partial pressure of SO2 in the balance. For each partial pressure, the amount of SO2 adsorbed was measured assuming that the preadsorbed water amount remained constant (no desorption of water).
Differential calorimetry coupled with manometry.
Adsorption heat of SO2 on dry cork and hydrated cork was measured with a differential calorimeter (Thian-Calvet Setaram C80, Setaram, Lyon, France) coupled with manometry. This device has been described in detail in other studies (Moise and Bellat 2005, Simonot-Grange et al. 1997, Weber et al. 2008), and it measures the molar adsorption enthalpy ΔHa(ma) of water (also called adsorption heat Qa) as a function of the adsorbed amount ma. Adsorption was attained at 298 K for cork powder samples of ~640 mg, previously outgassed under vacuum at 10−5 hPa for 72 hr. Sulfur dioxide pressure ranged from 10−5 to 40 hPa. For hydrated cork, the same protocol of hydration as previously described for thermogravimetry was followed. After outgassing under vacuum at 298 K, cork was submitted to a saturated water vapor pressure of 23 hPa. The water vapor adsorption equilibrium was reached after 12 hr, after which SO2 adsorption was performed on the hydrated cork sample to measure the adsorption heats.
Results and Discussion
Physicochemical characterization.
The physical structure of cork material is oriented in three directions: axial (vertical, parallel to the center of the tree), radial (horizontal), and tangential (perpendicular to the axial-radial plane). Cork stoppers are punched out along the axial dimension (Figure 1⇓). Cork pieces used in this study corresponded primarily to plane A (axial), which is in contact with the wine (horizontal storage) or the headspace of the wine bottle (vertical storage). According to SEM observations, the radial structure of cork is a homogeneous tissue composed of hexagonal thin-walled alveoli arranged in a honeycomb-type pattern with no intercellular spaces (Figure 1⇓). In axial and tangential planes, the cells appear as rectangular prisms, stacked base to base, parallel to the radial axis. The mechanical characteristics of cork are roughly isotropic in the plane perpendicular to the radial axis, as dictated by its special shape and cell structure (Gibson et al. 1981). However, these characteristics are anisotropic in both other planes, as also revealed by compression studies (Rosa and Fortes 1988). Average cork cells are 45-μm tall with a 20-μm hexagonal face and a 20-μm thickness (Gibson et al. 1981). Cork contains varying numbers of lenticular channels running radially (defining the quality grading), which are hollow and approximately cylindrical and constitute the macroscopic porosity (>50 nm diam, according to IUPAC; Sing et al. 1985). The existence of a small microporosity in cork (<2 nm diam, according to IUPAC; Sing et al. 1985), with a micropore volumn of 0.026 cm3·g−1, has been noted (Hanzlik et al. 2004) by adsorption of carbon dioxide in the pressure range 100 to 1500 hPa at a temperature of 298 K. However, for our sample, neither nitrogen adsorption (1 to 1000 hPa at 77 K) nor krypton adsorption (1 to 3 hPa at 77 K) was detectable. This lack of adsorption could indicate that there is probably no microporosity in the material or that pores are not accessible for nitrogen and krypton molecules if cell pore contraction occurs under vacuum outgassing and low temperature. From nitrogen adsorption manometry, the density of cork can be estimated as ~125 kg·m−3, in agreement with a very low-density material.
Adsorption isotherm of SO2 by dry cork.
The adsorption-desorption isotherm of SO2 by dry cork (plate geometry) at 298 K is shown (Figure 2⇓). The adsorbed amount is given in grams per 100 g (wt%) of cork activated under vacuum. The first adsorption branch was obtained from the outgassed sample and the second desorption branch from the sample at equilibrium under 40 hPa of SO2.
For this pressure range, which corresponds only to very low relative pressures of SO2 (0 <p/ps < 0.01), the adsorption capacity of cork is low (1.2 wt% under 40 hPa). This adsorption is extremely low compared with zeolite, which can reach 20 wt% (Marcu and Sandulescu 2004). Approximately 0.5% SO2 can be adsorbed under 5 hPa (Figure 2⇑). This pressure corresponds to the partial pressure of SO2 in the headspace of a bottle of wine, calculated with Henry’s law by taking, for the concentration of SO2 in wine, the maximal value of 400 mg·L−1. When extrapolated to the weight of a standard natural cork closure, the mass of SO2 adsorbed on cork stopper is equal to 15 mg, which represents <5% of the total SO2 content in a bottle of wine. Therefore, the amount of SO2 trapped by a cork stopper is negligible.
The Henry constant of adsorption, determined from the slope of the adsorption branch at low pressure, is equal to 1.6 x 10−3 g·g−1·hPa−1. This value is much lower than that of zeolite (0.03 g·g−1·hPa−1), a microporous adsorbent with a rather strong adsorption affinity of SO2 owing to the specific interactions between the polar sulfur molecules and the Lewis and Brönsted acidic sites at the surface of this material (Marcu and Sandulescu 2004). But this value is of the same order of magnitude as that one found on activated carbon (4 x 10−3 g·g−1·hPa−1) (Arunov et al. 1977). This weaker affinity of cork for SO2 is related to the chemical properties of its surface. Cork is composed of organic matter with few strong adsorption sites as localized positive charges (cationic sites) or acidic sites (hydroxyl groups) present in zeolites. It is thus expected that SO2 is essentially adsorbed on cork under the effect of dispersion forces. However, as outlined above, the adsorption isotherm of SO2 on outgassed cork exhibits a hysteresis loop between adsorption and desorption. Desorption levels for a given SO2 pressure are systematically slightly higher than adsorption levels. Such hysteresis phenomena have already been observed for other adsorbate/adsorbent systems, like n-butane/activated carbons or NH3, CH3OH, pyridine/montmorillonite (Gregg and Sing 1982), and can be attributed to a swelling of the material. For cork, this adsorption could similarly be due to the deformation of the structure of this nonrigid material during the adsorption process of the initial SO2 molecules, which consequently renders accessible other adsorption sites for the remaining SO2 molecules. During desorption process, these remaining SO2 molecules could be entrapped in the cork structure, when the initial adsorbed molecules are released. Nevertheless, the SO2 desorption process is not complete (Figure 2⇑); as some SO2 remains adsorbed on cork (~0.35%), even after pumping under vacuum for several days. A reactive adsorption of a small amount of SO2 with the cork surface is therefore suspected. Chemisorption also creates hysteresis on adsorption isotherm. As the adsorbed amount of SO2 is extremely low, the effect of swelling is likely negligible. So, the hysteresis phenomenon observed on the isotherm is due to chemisorption rather than to swelling.
The adsorption-desorption isotherm of SO2 by dry cork was also performed on a powder sample (~25 mg) (Figure 3⇓). Compared with the previous adsorption isotherm on the same material in plate geometry, the amounts of SO2 adsorbed at equilibrium are systematically higher in the whole pressure range. Nevertheless, the adsorption capacity still remains at a very low level (maximum of ~1.7 wt% for 40 hPa). The isotherm displays similar adsorption and desorption branches, with the same hysteresis phenomenon and a constant nondesorbed fraction of SO2 molecules. Cork powder has a particle size <500 μm and therefore exhibits a larger developed surface area for SO2 adsorption, which could favor a surface adsorption mechanism. The adsorption mechanism of SO2 on cork is similar to that on nonporous or macroporous solids.
Adsorption heat of SO2.
The dependence of the loading on the adsorption heats of SO2 (given in absolute value) on dry cork is shown (Figure 4⇓). For the first adsorption, the heat of adsorption at low filling was ~100 kJ·mol−1, an unusually high value for an adsorption process that is supposed to occur on a hydrophobic plane surface (Gomes et al. 1993). The adsorption heat sharply decreased as the filling increased. Above 1 wt% of adsorbed SO2, it finally tends toward the enthalpy of liquefaction of SO2, which is equal to 22.9 kJ·mol−1 (Lide 2005a), as expected for a physisorption process. Such high adsorption heats at low filling have already been reported for SO2: ~63 kJ·mol-1 for adsorption on graphitized carbon blacks near 0 K (Beebe and Dell 1955), ~100 kJ·mol−1 for adsorption on carbon blacks at 323 K (Murphy et al. 1977), and up to ~500 kJ·mol−1 for adsorption on chrysotile at 323 K (Murphy and Ross 1977). These high energies are typical of chemical bonds. A second adsorption was also performed after outgassing the sample under vacuum at 298 K for 72 hr in order to evacuate all the molecules physisorbed at the surface. In this case, the adsorption heat at low filling was reduced by half, ~50 kJ·mol−1, a value usually observed in physical adsorption. These results show that the first SO2 molecules in contact with cork strongly reacted with the surface and suggest a chemisorption mechanism for the first adsorbed molecules. Thus, it appears that a binding of a small fraction of SO2 molecules involved high-energy binding with a few specific adsorption sites on the cork surface. At this stage of our study, it is not possible to identify the adsorption sites at the cork surface. Other studies indicate that the physisorption sites could be carboxylic groups or ending OH groups (Chubar et al. 2003, Garcia-Martinez et al. 2002). For the reactive sites, a small fraction of SO2 could also react with the hydroxyls to form HSO3- as chemisorbed specie. Alternatively, SO2 could be bound to the organic surface through C-O and/or C-S bonds, as has been observed on graphite surface (Pliego et al. 2005).
Molar entropy of the adsorbed phase.
The molar entropy of the adsorbate is an interesting value that gives information about the physical state of the matter adsorbed at the surface. In the physisorption equilibrium gas ↔ adsorbate, the molar entropy of the adsorbed phase is given by the following relation:
(1)
with: Sa(ma) = molar entropy of the adsorbed phase at a given loading ma (J·mol−1·K−1)
ΔSa(ma) = molar entropy of adsorption (J·mol−1·K−1)
S(g)0 = gas standard molar entropy = 248 J·mol−1·K−1 for SO2 at 298 K (Lide 2005b)
ΔSa(ma) is determined from the calorimetric adsorption heat (equivalent to the isosteric heat in our experimental conditions) and the adsorption isotherm with the following equation (Bellat and Simonot-Grange 1995) at constant ma:
(2)
with: ΔHa(ma) = molar enthalpy of adsorption (J·mol−1)
T = adsorption temperature (K)
R = ideal gas constant (8.31 J·mol−1·K−1)
p = equilibrium pressure at the loading ma (Pa)
p0 = standard pressure (105 Pa)
The evolution of the molar entropy (Sa) of SO2 adsorbed on cork as a function of the loading at 298 K is shown (Figure 5⇓). Standard molar entropy values for gas, liquid, and solid SO2 at 298 K were obtained using PhysProps software (G&P Engineering, League City, TX). The results obtained from the first adsorption show very low molar entropy of SO2 at low filling, close to zero. Such values are much too low for a sole physisorption phenomenon. For physisorption, molar entropy usually tends at minimum to the solid entropy when the adsorbate is highly confined, as in the adsorption of mercaptan or saturated and aromatic hydrocarbons in zeolites (Bellat and Simonot-Grange 1995, Weber et al. 2008). This could be explained by chemisorption of the initial SO2 adsorbed molecules on cork, in which case the molar entropy calculated by Equation 2 has no physical meaning, because a chemisorption phenomenon adds to SO2 physisorption. On the other hand, the molar entropy of SO2 calculated for the second adsorption ranges between that of solid and gas, corresponding to a physisorption process. At low filling, the molar entropy is close to that of the solid. The SO2 molecules are “frozen” on their adsorption sites. At higher filling, the molar entropy lies between that of liquid and gas, indicating that the SO2 molecules have higher degrees of freedom at the surface. If at low covering the adsorption process seems rather localized, it is replaced by a mobile adsorption process at higher covering.
Effect of cork hydration on SO2 adsorption.
Cork, in its common use of closure, is in contact not only with SO2 but also with many volatile organic molecules. The head-space of a wine bottle contains mainly water vapor, but it also contains other organic vapors such as ethanol or aroma compounds in equilibrium with the liquid phase. Therefore, the adsorption process of gaseous SO2 can also depend on other chemical species adsorption. We focus here on the effect of the presence of water previously adsorbed on cork.
The adsorption isotherm of SO2 on cork containing 5 wt% of water is shown (Figure 6⇓), with that on dry cork reported for comparison. The isotherm was significantly modified when SO2 was adsorbed on hydrated cork, with the adsorption capacity divided by a factor of 3, which at first may seem surprising as SO2 is highly soluble in water (www.henrys-law.org). Indeed, given its high solubility, it was expected that SO2 would be adsorbed in higher quantities when cork was hydrated. Moreover, adsorbed water molecules could be active sites for SO2 adsorption by formation of hydrogen bonds with SO2 molecules, thus increasing the ability to adsorb this soluble gas. In the absence of oxygen, any reaction of oxidation of SO2 as SO3 and hydration of SO3 as H2SO4, as observed on activated carbons, can be excluded (Arunov et al. 1977, Bagreev et al. 2002, Furmaniak et al. 2006, Kisamori et al. 1994, Zhang et al. 2007). Therefore, we suspect that the decrease of the adsorption capacity of SO2 in the presence of water was the result of a competitive adsorption between water and SO2. After preadsorption of water, many sites could not directly interact with SO2, as they were already occupied by water molecules. Our calorimetric measurements seem to support this hypothesis. The adsorption heat of SO2 on hydrated cork powder, measured in a larger pressure range than for dry cork (up to 60 hPa), is shown (Figure 7⇓). The shape of the calorimetric curve was very different in the presence of preadsorbed water. At zero loading, the adsorption heat is much lower than that observed on dry cork (60 vs 100 kJ·mol−1). The adsorption heat increased with the loading to reach a maximal value of ~110 kJ·mol−1, before decreasing sharply to ~50 kJ·mol−1. The maximal value of the adsorption heat is observed at the filling of 0.35 wt%, which is exactly the value of the amount of SO2 chemisorbed on the surface, as observed by thermogravimetry (Figure 2⇑). It is thus tempting to attribute the ascending section of the calorimetric curve to the chemisorption of SO2 on sites previously occupied by water molecules. The heat involved would correspond on one hand to the exothermic chemisorption of SO2 and on the other to the endothermic desorption of water, thus explaining why the resulting adsorption heats are lower than those measured on dry cork. It is difficult to clarify why the adsorption heat increases in this range of loading. It may be due to adsorbate-adsorbate interactions, particularly between water and SO2, which are certainly engaged in hydrogen bonds. The descending section of the calorimetric curve would correspond to the physisorption of SO2 on sites likely free of water once the chemisorption sites were completely saturated. However, this interpretation remains a hypothesis, with no solid experimental evidence for confirmation, and it must be considered with caution.
Conclusion
For the first time the adsorption of sulfur dioxide on cork has been studied in the gas phase by thermogravimetry and calorimetry. The adsorption capacity of SO2 has been determined with accuracy at 298 K and for pressures up to 40 hPa. The adsorption capacity of SO2 on cork is rather low. Thus, it has been possible to show that the amount of SO2 that can be trapped by a standard cork stopper from the headspace of the bottle wine is negligible. Therefore, thermodynamics of adsorption on cork cannot account for the decrease of SO2 content in wine during aging. We indicate on the adsorption isotherm a hysteresis loop in desorption that has been attributed to a reactive adsorption of SO2 on cork. A hysteresis in desorption due to a swelling of the material is not excluded but is unlikely considering the low adsorbed amount. Nevertheless, the amount of SO2 chemisorbed on the surface is low, and the interaction of SO2 with cork is mainly governed by a physisorption process. Surprisingly, in the presence of water preadsorbed on cork, the adsorption capacity of SO2 is divided by a factor of 3. This decrease in the adsorption capacity would result from the occupancy of the adsorption sites by water molecules. The only sites accessible to SO2 would be those where SO2 is chemisorbed. It is only on these strong specific sites that SO2 could find the energy required for displacing the water previously adsorbed on them. Thus, competitive adsorption mechanisms between water and SO2 could occur on cork, also indicating that many other polar molecules, such as ethanol, are susceptible to adsorb selectively with SO2 on cork. In that case, the real adsorption capacities could be very different than those determined from the adsorption of single components on cork. Consequently, thermodynamics of coadsorption of SO2 with other molecules, in particular with water, must be further investigated. Moreover, the adsorption of SO2 has been studied only on crude cork. It will be of a great interest to study the effects of various chemical treatments of cork, such as washing and paraffining, on its adsorption properties. Study of adsorption and coadsorption of various compounds on crude and pretreated cork is in progress.
Footnotes
Acknowledgments: The authors gratefully acknowledge the Bureau Interprofessionnel des Vins de Bourgogne and the Regional Council of Burgundy for financial support of this work (project 2008-9201AAO024S00120).
The authors thank Christian Paulin and Claudie Josse, Institut Carnot de Bourgogne, for assistance with thermogravimetry and SEM, respectively.
- Received November 2008.
- Revision received January 2009.
- Accepted January 2009.
- Published online June 2009
- Copyright © 2009 by the American Society for Enology and Viticulture