The Scientific World Journal

Volume 2012, Article ID 249041, 7 pages

doi:10.1100/2012/249041
The cientificWorldJOURNAL

Research Article
Acid-Base Chemistry of White Wine: Analytical Characterisation
and Chemical Modelling

Enrico Prenesti,1 Silvia Berto,1 Simona Toso,2 and Pier Giuseppe Daniele1
1 Dipartimento di Chimica Analitica dell’Università degli Studi di Torino, Via Pietro Giuria 5, 10125 Torino, Italy
2 BP Italia S.p.A. Divisione Castrol Industrial, Environment Park, Via Livorno 60, 10144 Torino, Italy

Correspondence should be addressed to Silvia Berto, [email protected]

Received 28 October 2011; Accepted 8 December 2011

Academic Editors: E. Garcia-Moruno and J. Gebler

Copyright © 2012 Enrico Prenesti et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

A chemical model of the acid-base properties is optimized for each white wine under study, together with the calculation of their
ionic strength, taking into account the contributions of all significant ionic species (strong electrolytes and weak one sensitive to
the chemical equilibria). Coupling the HPLC-IEC and HPLC-RP methods, we are able to quantify up to 12 carboxylic acids, the
most relevant substances responsible of the acid-base equilibria of wine. The analytical concentration of carboxylic acids and of
other acid-base active substances was used as input, with the total acidity, for the chemical modelling step of the study based on
the contemporary treatment of overlapped protonation equilibria. New protonation constants were refined (L-lactic and succinic
acids) with respect to our previous investigation on red wines. Attention was paid for mixed solvent (ethanol-water mixture), ionic
strength, and temperature to ensure a thermodynamic level to the study. Validation of the chemical model optimized is achieved
by way of conductometric measurements and using a synthetic “wine” especially adapted for testing.

1. Introduction In our previous paper on red wines [2] we needed to

quantify all acid-base active substances in order to investigate
Nowadays the wine has achieved considerable importance in the acid-base chemistry of red wine at a speciation level by
the global economy. The chemical characterization aimed to way of a careful equilibrium analysis. A chemical model was
deliver an excellent product to the consumer assumes a grow- proposed, based on the simulation of the alkalimetric titra-
ing importance in the world of oenological analysis. Par- tion curve of each wine. The simulated titration curve was
ticularly, the acidity of wine is a characteristic of central obtained by way of a multicomponent equilibrium calcula-
importance. Acid-base chemistry of wines is mainly based on tion made thanks to suitable values of protonation constants
the protonation/deprotonation status of the carboxylic acids. of significant pH-determining substances found in wines.
Carboxylic acids belong to a class of natural compounds An HPLC-RP separation of carboxylic acids and proline was
widespread distributed in fruits and vegetables. Moreover, executed, improving the method of Tusseau and Benoit [3].
they are extensively used as food acidulants in the manufac- Unfortunately, a coelution of acetic and 2-ketoglutaric acids
turing of beverages, fruit- and vegetable-based drinks or was observed under the adopted experimental conditions. To
juices, and as antioxidants, acidifiers and drug adsorption ensure the separation of these two acids, we have now used an
modifiers in pharmaceutical industries. Physical and chemi- ion-exclusion chromatographic (HPLC/IEC) separation [4,
cal analyses of wines have become one of the most important 5] with photometric detection. We propose improvements
aspects of the modern quality control in the food field. Par- in both analytical and equilibrium aspects of our study. Cou-
ticularly, in enology the nature and concentration of the car- pling the current HPLC-IEC method with previous HPLC-
boxylic acids are very important for (i) acid-base and redox RP one, we are now able to quantify up to 12 carboxylic acids.
properties, (ii) biological and colour stability, (iii) organ- Further protonation constants, with respect to those previ-
oleptic characteristics, and (iv) monitoring of fraudulent ously determined [2, 6], were now refined (at 25◦ C in mixed
practices and/or alterations [1]. ethanol-water solvent in KCl 0.05 or 0.1 M)—L-lactic and
2 The Scientific World Journal

succinic acids—to strengthen the thermodynamic approach a 20 μL sampling loop, and the detector was an Agilent 1100
to the acid-base chemistry of wines. The analytical concen- UV-vis photometer.
tration of carboxylic acids, and other inorganic and organic As to HPLC-RP analyses, the chromatographic separa-
acid-base active substances, was used as input, with the total tions were performed on a Merck Superspher 100 RP-18 end-
acidity, for the chemical modelling step of the study, assisted capped (250 × 4 mm I.D.) spherical phase column with a
by a specific software based on the contemporary treatment particle size of 4 μm.
of overlapped protonation equilibria. Attention was paid for The HPLC-IEC analyses were carried out using a Supel-
mixed solvent (ethanol-water mixture), ionic strength, and cogel C-610H column (300 × 7.8 mm I.D.), polystyrenediv-
temperature, to ensure a thermodynamic level to the study. inylbenzene-based strong acid cation exchange resin in the
Six white wines were included in this investigation: (i) H+ -form (9 μm particle size).
four white wines are produced in Piedmont (North-West
Italy), named Erbaluce and Cortese; two vintages were con- 2.4. Chromatographic Conditions for Carboxylic Acids Analy-
sidered for these wines, that is, Erbaluce 2007 and 2008 sis. As for HPLC-RP separation, the mobile phase was com-
(henceforth E07 and E08) and Cortese 2007 and 2008 posed of 70 g/L (0.52 M) potassium dihydrogen phosphate
(henceforth C07 and C08); (ii) one Californian wine: and 14 g/L (0.10 M) ammonium sulphate adjusted to pH 2.1
Chardonnay 2009, a white wine produced in Sonoma Valley with phosphoric acid, in order to have the highest proto-
(San Francisco) (henceforth Ch09); (iii) one French white nation degree of the acids under examination, according to
wine, produced in Alsace, namely, Riesling 2007 (henceforth the method of Tusseau and Benoit [3]. The flow rate was
R07). 0.7 mL/min at room temperature. The separation was carried
A chemical model is optimized for each white wine out under isocratic conditions, and the detection was effected
together with the calculation of its ionic strength taking by measurement of the UV absorption at 210 nm. See further
into account the contributions of all significant ionic species details in [2].
(strong electrolytes and weak one sensitive to the chemical For HPLC-IEC separation, the mobile phase was com-
equilibria). Validation of the chemical model is achieved by posed of 10% acetone and 0.5 mM sulphuric acid. The flow
way of conductometric measurements and using a synthetic rate was 0.4 mL/min. Column and precolumn were thermo-
“wine” especially adapted for testing. stated at 45◦ C. The separation was carried out under isocratic
conditions, and the detection was effected by measurement
2. Experimental of the UV absorption at 210 nm.
Each substance determined via HPLC was identified
2.1. Sample Storage. All wines were stored at room tempera- by its retention time in comparison with the response of
ture in the dark and were subdivided into small glass bot- standard solution of pure compounds. Standard addition of
tles (250 mL capacity) to avoid air contact and other con- some substance to the wine was performed in order to verify
tamination. the attribution of the peaks.

2.2. Reagents and Solutions. Standard solutions of carboxylic 2.5. Preliminary Treatments of Wine Samples for Carboxylic
acids were prepared from analytical-reagent grade chemicals; Acids Analysis. Both for HPLC-RP and HPLC-IEC the sam-
standardization process was performed to achieve high- ples were treated by means of Chromabond C18 (500 mg)
quality, by titrating each stock solution with standard car- MACHEREY-NAGEL cartridges, which allow to purify the
bonate-free KOH. matrix of those molecular compounds (polyphenols and sac-
Carboxylic acids: 2-ketoglutaric, shikimic, L-lactic, L-cit- charides, as an example) which can interfere in the chro-
ramalic, and gallic acids were from Sigma; L-tartaric and suc- matographic measurements. This treatment does not alter
cinic acids were from Carlo Erba; DL-malic, acetic, and citric the carboxylic acid composition of the samples, as verified by
acids were from Merck; pyruvic and fumaric acids were means of a check on synthetic mixtures. Each cartridge was
from Fluka. Amino acid: L-proline was from Sigma. Salts: conditioned with little volume (few mL) of water and then
KH2 PO4 was from Sigma, (NH4 )2 SO4 was from Carlo Erba. with little volume (few mL) of ethanol, before sample purifi-
Solvents: acetone was from Labochem, ethanol (96% w/v) cation procedure. As for ion-exclusion chromatography, each
was from Merck. Mineral acids: phosphoric acid (85% w/v) sample was also purified from metal ions in order to prevent
was from Carlo Erba, sulfuric acid (95–97% w/v) was from the saturation of the cation exchange resin surface of the
Fluka. Standard solutions of KOH were prepared by diluting column used for the separation. 1 g of a cation exchange resin
concentrated Merck ampoules and were standardised against (Amberlite, IR-124) was added to 100 mL of each wine and
potassium hydrogen phthalate (Fluka, puriss.). Amberlite stirred for twenty minutes by means of a magnetic stirrer.
(IR-124), a cation exchange resin, was from Carlo Erba. Finally, each sample was filtered through a 0.45 μm Millipore
Grade A glassware and deionised and twice distilled water filter and diluted 1/10 (v/v) with twice distilled water before
were used for all solutions. the chromatographic injection.

2.3. Chromatographic Apparatus for Carboxylic Acids Analysis. 2.6. Ion Chromatography for Inorganic Anions Analysis. As
Chromatographic analyses for carboxylic acids were carried for inorganic anions, analysis was carried with a Metrohm
out with an Agilent 1100 chromatograph equipped with an 690 Ion Chromatograph equipped with a Bischoff HPLC
Agilent 1100 pump. The injector was a rheodyne valve with pump. The chromatographic separations were performed on
The Scientific World Journal 3

a PRPTM-X100 column (polystyrenedivinylbenzene-trimeth- 2.12. Alcohol Content. The alcohol (ethanol) contained in
ylammonium exchanger spherical phase column). The sep- each wine (symbol “% vol.”) was determined, according to
aration was performed at pH ≈ 8. The mobile phase was the official method [8], by means of distillation and further
composed of 4 mM p-hydroxybenzoic acid and 1% methanol density measurement (20◦ C).
adjusted to pH 8.5 with NaOH (in order to have maximal
deprotonation of the analytes). The flow rate was 1.5 mL/min 3. Results and Discussion
at room temperature.
3.1. Analytical Determinations
2.7. ICP/AES. A model Liberty 2 ICP/AES was used to deter-
mine metal ions in wine. Potassium, sodium, calcium, mag- 3.1.1. Quantification of Organic Acids. First, to quantify car-
nesium, and iron were determined after suitable dilution of boxylic acids in wine, we chose to adopt a method without
each sample, depending upon the nature of the analyte. The derivatization based on a classical RP separation (as in
quantification was based on external standards. our previous paper [2]). The conditions adopted allowed
us the separation of the following acids (in parenthe-
2.8. Potentiometric Apparatus. Potentiometric measure- ses, the retention time is indicated): tartaric (3.6 min),
ments were performed at T = 25 ± 0.1◦ C and ionic strength malic (5.0 min), shikimic (5.8 min), lactic (6.7 min), citric
I = 0.05 or 0.1 M (KCl) with C2 H5 OH (EtOH) at 12% level (11.9 min), fumaric (12.8 min), succinic (13.7 min), citra-
with a model 713 Metrohm potentiometer equipped with malic (15.7 min), and gallic (27.9 min), together with the
a combined glass electrode. The titrant was dispensed with simultaneous determination of the proline (4.3 min), the
a model 765 Dosimat burette by Metrohm. The electrode most abundant amino acid present in wine [1]. In Figure 1(a)
couple was calibrated in −log[H+ ] units (pH) employing a standard run is shown. Calibration graphs (peak height or
alkalimetric titrations of hydrochloric acid with standard, area versus substances concentration) were built in order to
carbonate-free, potassium hydroxide. Ionic strength, ionic characterize the acidic profile of wines under investigation,
medium, and ethanol percentage of the calibrating solu- starting by a chromatographic run of a stock solution and
tions were the same as the solutions being examined. The its further dilutions. In Figure 1(b) the chromatographic
alkalimetric titrations were carried out in a stream of puri- separation on Ch09 is given as an example. Unfortunately,
fied nitrogen gently bubbled in the titration cell. Tempera- a coelution of acetic and 2-ketoglutaric acids is observed
ture control was achieved by means of a circulation of water, under the adopted experimental conditions at a tR of 7.2
in the outer chamber of the titration cell, from a model minutes. Plot of the integrated peak area and/or height
D1-G Haake thermocryostat. Each titration was at least re- against concentration (mg/L) of each molecule was always
peated three time. linear (the correlation coefficient R ranged between 0.9997
and 0.9999) in the concentration ranges investigated.
2.9. Preliminary Treatments for Titrations. Before the acid- With HPLC-IEC the conditions adopted allowed us the
base titration each wine was filtered through a 0.45 μm Mil- separation of (in parentheses, the retention time is indicated)
lipore filter; CO2 was then removed by means of strong stir- 2-ketoglutaric (10.8 min), pyruvic (11.8 min), and acetic
ring under vacuum (few minutes), according to the indica- (23.2 min). Calibration graphs were obtained by plotting
tions of official methods [8]. peak height or area against analytes concentration and excel-
lent linearity was found for each standard molecule, being
2.10. Equilibrium Calculations. The nonlinear least squares R in the range 0.9992 ≤ R ≤ 1. The repeatability of peak
computer program ESAB2M was used to evaluate the purity heights or areas, obtained by repeated (at least three) injec-
of the reagents (starting from acid-base titration data) and tions (HPLC-IEC) of the same concentration of standard, is
to refine all the parameters related to the calibration of the in the range 0.6–4% (relative standard deviation) for all the
electrode system [9]. carboxylic acids studied. Intraday and infraday repeatabili-
The protonation constant values were expressed by the ty was of very similar extent.
general formula: β pq = [L p Hq ]/[L] p [H]q and refined (as In Table 1, the concentration (mg/L) of carboxylic acids
log βH ) by means of the BSTAC program, which minimises utilized for the optimization of the chemical model is col-
the error squares sum on electromotive force values and is lected for each wine under investigation. Comparing the re-
able to take into account (if desired) eventual variations of sults with those obtained on red wines analysed in the pre-
ionic strength among and/or during titrations [10]. vious paper [2], in white wines a higher concentration of cit-
Distribution diagrams and simulated titration curves ric acid was measured while citramalic and gallic acids are
were obtained using the computer program ES4ECI [10]. found only in the red wines previously analysed.
The slight contraction in volume around 0.3–0.4% [2],
corresponding to the ethanol-water mixture formation, was 3.1.2. Quantification of Inorganic Anions and Cations. Results
considered in all measurements and calculations. for inorganic anions and cations are collected in Tables 2 and
3 respectively.
2.11. Conductometric Apparatus. A model 160 Amel con- The inorganic anions were determined with HPLC-IC,
ducimeter equipped with a model 196 Amel electrode, was and the conditions adopted allowed us the separation of
used as indicating device during alkalimetric titration of (in parentheses, the retention time is indicated) phosphate
wines (T = 25◦ C). (7.1 min), nitrate (6.3 min), sulphate (3.9 min), and chloride
4 The Scientific World Journal

Sk
M T
100 100
80 F 80
(mAU)

(mAU)
60 L 60 M
40 C Pro Sk
A+K 40
Pro G
20 S Cm L A+ K
20 C S
0 0
0 5 10 15 20 25 2.5 5 7.5 10 12.5 15 17.5 20 22.5
(min) (min)
(a) (b)

Figure 1: HPLC-RP separation: (a) chromatogram of a standard mixture of carboxylic acids. Concentrations in mg/L are T 100, Pro 250, M
250, Sk 5, L 250, A 150, K 100, C 250, F 3, S 550, Cm 300, G 2.5. (b) Chromatogram of Ch09 (dilution 1/10 v/v). Run stopped at 22.5 min.

Table 1: Analytical concentrations (mg/L) of carboxylic acids and proline quantified in each wine under investigation by HPLC-RP and
HPLC-IEC. The relative standard deviation, evaluated on three replicates on each sample, ranges between 0.6 and 10%.

Wine
Substances Identification Symbol
E07 E08 C07 C08 Ch09 R07
Acetic acid A 395.3(a) 349.1(a) 710.1(a) 528.1(a) 366.5(a) 240.3(a)
Citramalic acid Cm <d.l.(b) <d.l. <d.l. <d.l. <d.l. <d.l.
Citric acid C 718.5 257.4 499.5 426.5 480.3 326.6
Fumaric acid F <d.l. <d.l. <d.l. <d.l. <d.l. <d.l.
Gallic acid G <d.l. <d.l. <d.l. <d.l. <d.l. <d.l.
2-Ketoglutaric acid K 58.4(a) 190.0(a) 24.8(a) 33.6(a) 26.3(a) 23.4(a)
Lactic acid L 614.5 653.8 1116.6 1000.4 758.5 374.4
Malic acid M 2419.0 3038.5 449.2 682.5 1702.9 2373.4
Proline Pro 300.4 308.5 545.6 527.2 4109.1 1922.2
Pyruvic acid Py 201.5(a) 236.7(a) 214.7(a) 196.2(a) 44.9(a) 13.2(a)
Shikimic acid Sk 34.8 33.1 20.9 13.9 34.8 52.3
Succinic acid S 916.4 866.8 395.6 560.9 259.8 236.2
Tartaric acid T 1805.6 2053.2 1900.1 1678.0 2926.8 4247.5
(a)
Data obtained from HPLC-IEC technique.
(b) d.l. = detection limit.

Table 2: Concentration (mM) of inorganic anions in each wine. The uncertainty (three replicates) ranges between 2 and 10% (±s).

Wine
Inorganic anion
C07 C08 E07 E08 Ch09 R07
Phosphate 3.77 3.95 4.18 4.47 5.22 3.28
Nitrate <d.l.(a) <d.l. <d.l. <d.l. 0.26 <d.l.
Sulphate 14.04 12.48 17.49 17.88 4.83 3.46
Chloride <d.l. <d.l. <d.l. <d.l. <d.l. <d.l.
(a)
d.l. = detection limit.

Table 3: Concentration (mM) of the metal ions in each wine (ICP/AES). The uncertainty (three replicates) ranges between 2 and 8% (±s).

Wine
Metal ion
C07 C08 E07 E08 Ch09 R07
Ca 2.2 2.2 2.5 2.5 1.64 2.70
Fe 0.04 0.03 0.09 0.05 0.04 0.04
K 13.2 11.5 22.8 16.9 12.93 17.96
Mg 3.1 2.9 3.4 3.6 4.27 3.05
Na 1.3 1.3 0.7 0.5 1.33 0.75
The Scientific World Journal 5

Table 4: Overall protonation constant values, as log βiH , at two ionic strength values (0.05 and 0.1 M), 0% and 12% of ethanol, K+ Cl− as
background salt, T = 25◦ C. The uncertainty is reported in parentheses as standard deviation in the last significant digit.

log βiH (KCl 0.05 M) log βiH (KCl 0.1 M)

Substance i
EtOH 0% EtOH 12% EtOH 0% EtOH 12%
L-Lactate 1 3.70(a) 3.870 (3) 3.81(b) 3.66(c) 3.830 (3) 3.77(b)
1 5.319(a) 5.49 (2) 5.467(b) 5.24(c) 5.39 (2) 5.38(b)
Succinate
2 9.347(a) 9.68 (5) 9.651(b) 9.23(c) 9.59 (3) 9.53(b)
(a)
Values at I = 0.05 M are calculated by way of a Debye-Hückel-type equation from literature data [2].
(b) Data from [2].
(c) Data from [7].

Table 5: Alcohol content (% vol.) and acid-base results of each wine.

Wine % vol. pH Total acidity as CH (mM) Total acidity

Potentiometric detection Conductometric detection as g/L Tartaric acid
C07 11.7 3.13 66.60 66.00 4.95
C08 11.8 3.12 65.20 65.20 4.89
E07 12.3 3.24 92.20 92.20 6.92
E08 12.3 3.05 96.00 96.00 7.20
Ch09 13.3 3.21 78.70 79.06 5.90
R07 12.3 3.16 86.40 86.80 6.48

(3.2 min). Calibration graphs were obtained by plotting peak correct. These experimental values of log βH in 12% ethanol-
area against analytes concentration, and excellent linearity water mixture were reported in Table 4 and were compared
was found for each standard molecule, being R in the range with those obtained with the calculation previous described
0.9993 ≤ R ≤ 0.9998. The results reported in Table 2 show [2]. The comparison is satisfactory. The log βH in aqueous
that sulphates are predominant, while nitrates were found in medium were also collected in Table 4. The values at I =
low concentration only in Ch09 and chlorides resulted under 0.1 M are from [7], whereas the values at I = 0.05 M are cal-
the detection limit. culated by way of a Debye-Hückel-type equation as reported
in [2]. Any chemical model developed in this paper is based
As to cations, the results reported in Table 3, obtained by
on the set of protonation constants presented in [2] with the
ICP/AES, show that potassium is the predominant one in all
integration currently obtained for L-lactic and succinic acids
wines.
(Table 4).
3.2. Equilibrium Determinations 3.2.2. Alkalimetric Titration of Wines. Table 5 reports pH,
alcoholic grade and total acidity for each wine. CO2 was pre-
3.2.1. Protonation Constants in Mixed Solvent. In our pre- liminarily removed by means of brief stirring under vacuum.
vious paper [2] three carboxylic acids—acetic, L-tartaric, The pH was measured on undiluted wines, at T = 25◦ C, and
and citric acid—chosen as model substances, were titrated expressed as −log[H+ ]. Wine was diluted 1 : 20 (v/v) for total
in order to refine the overall protonation constant values acidity measurement. During the alkalimetric titration of
(log βH ) for mono-, di-, and triprotic acids in ethanol- each wine, we found via potentiometric detection (combined
water media. We assumed that the differences experimentally glass electrode) the first inflection point at about pH 7.5,
observed (Δ log βH ) for each of our three model molecules as expected. The strong base used up to this flex allows
in water and ethanol-water media are the same under the the calculation of the total acidity parameter CH (also ex-
same conditions of electrical charges involved in any single- pressed as g/L of tartaric acid), fundamental in the chemical
step protonation reaction. The terms Δ log βH were applied modelling step of this work.
to calculate the values of log βH for the substances not titrat- Conductometric data were also recorded, during the
ed, starting by the values in aqueous solution. Now, we add- alkalimetric titrations of each wine. As in potentiometry, the
ed L-lactic and succinic acids to improve the model by verify- derivative graph can be used to estimate the end point in the
ing previous assumptions with common substances present conductivity versus titrant volume curve. The estimations of
in wine but lightly different from those previously consid- the CH obtained by potentiometry and conductometry are in
ered. Conditions: KCl at two ionic strength values, I = 0.05 excellent agreement (Table 5).
and I = 0.1 M, was used as background electrolyte in 12%
ethanol-water mixture. Since K+ ion is the most represent- 3.3. Chemical Modelling
ative metal ion in wines [1], we believe the choice of KCl as
background salt to obtain a suitable set of protonation con- 3.3.1. Building of the Chemical Model. For each wine a chem-
stant values to model the acid-base chemistry of wines is ical model can be built taking into account the analytical
6 The Scientific World Journal

Table 6: Chemical modelling. Equilibrium-based simulation at 25◦ C to obtain a calculated value of pH (pHcalc ) according to the four
chemical models depending upon the ionic strength value and the percentage of ethanol.

pHcalc
Wine % vol. pHexp (a) Model 1 Model 2 Model 3 Model 4
EtOH 0%, KCl 0.05 M EtOH 0%, KCl 0.1 M EtOH 12%, KCl 0.05 M EtOH 12%, KCl 0.1 M
C07 11.7 3.13 2.85 2.85 2.99 2.99
C08 11.8 3.12 2.86 2.86 2.99 2.99
E07 12.3 3.24 2.98 2.98 3.13 3.13
E08 12.3 3.05 2.86 2.86 3.00 3.00
Ch09 13.3 3.21 3.06 3.06 3.19/3.21(b) 3.20
R07 12.3 3.16 3.07 3.07 3.22 3.22
(a)
pHexp is the experimental pH measured at T = 25◦ C and expressed as −log[H+ ].
(b) pH value corrected for % ethanol.

Table 7: Values of ionic strength calculated for each wine as output of Model 3.

Wine C07 C08 E07 E08 Ch09 R07

Icalc (M) 0.055 0.051 0.070 0.071 0.042 0.041

concentrations of any acid-base active substance analysed show the importance of the ethanol, while the role of the
and the refined values of the protonation constants in the concentration of the background salt, in the range investi-
suitable chemical medium. As to current wines, a greater gated, is of minor relevance.
amount of SO2 is found with respect to red wines previously For the Ch09, which has an alcohol content significantly
investigated [2]. Nevertheless, the most of SO2 is in the exceeding 12% vol., we took into account the influence of
combined form (as Bertagnini adduct), and it does not affect the amount of ethanol by applying the corrective procedure
the acid-base equilibria for pH < pHendpoint . Moreover, the previously described [2]. The method allows to avoid the use
contribution of the free SO2 (measured by titration with I2 , of the specific set of log βH values at the current alcoholic
according to the official methods [8]) resulted negligible con grade. We used the avalilable set of log βH at 8, 12, and 16%
sidering that the pKa2 = 6.85 of H2 SO3 (25◦ C, I = 0.1 M in of ethanol (I = 0.05 M, KCl) [2] to calculate three pH values
water) hinders its fully deprotonation at the experimental pH for each wine. A linear fitting was applied to the points of
value of the endpoint of the titration curve of wine. Similar the diagram % ethanol versus pH (see Figure 2), and the
behaviour is evidenced for phosphoric acid (pKa2 = 6.75, corrected value of pH was then interpolated. The correction
25◦ C, I = 0.1 M in water). for Ch09 provided pH = 3.21 starting by pHmodel3 = 3.19
(pHexp = 3.21).
13 acid-base active substances (in the field of pH <
We can estimate the accuracy by way of the difference
pHendpoint ) were considered as reactants during the input
between measured and calculated pH of each wine: |pHexp −
building for the computer-assisted simulation of the alka-
pHcalc |. The average value of six white wines is 0.05, a result
limetric titration of each wine, 5 metal ions (and nitrate
that can be considered satisfactory, particularly taking into
ions for Ch09) were considered as background electrolytes
account the intrinsic uncertainty of the pH reading (±0.02
(contributing to the ionic strength of each fluid), and 20
[11]) and the wide variety and amount of data—of both
protonation equilibria were treated at the same time. We
analytical and thermodynamic nature—used as input in the
considered the hydrolysis of iron(II) (wine is under reducing
simulation.
conditions), but the influence on pH calculation is not ap-
preciable. The model investigation was stopped at pH ≈ 6.5. 3.3.2. Validation of the Chemical Model. The output based on
The pH value of each wine was then calculated (pHcalc ) the chemistry simulated by Model 3 also contains the ionic
as a result of all the multiple chemical protonation equilibria strength (molar scale) calculated for each wine. As shown
set. Table 6 shows the pHcalc values for each wine obtained in Table 7, the mean value of the ionic strength is around
by various simulations. Four chemical models were tested 0.05 M (lightly higher for E07 and E08, near 0.07 M), as that
considering the following: (i) Model 1: the values of each estimated for the red wines previously studied [2]. The trend
log βH in water at I = 0.05 M (KCl), (ii) Model 2: the values of the ionic strength with varying the pH in the range 3–6
of each log βH in water at I = 0.1 M (KCl), (iii) Model 3: the follows that of the conductivity, indicating the consistency
values of each log βH at I = 0.05 M (KCl) in 12% ethanol, and of our assumptions and the applicability of the chemical
(iv) Model 4: the values of each log βH at I = 0.1 M (KCl) in model optimized. Conductometric and pH-metric outputs
12% ethanol. show fruitful convergence—experimentally proved by the
The best simulation of the acid-base chemistry of each measurement of the total acidity (Table 5)—and comparable
wine is obtained by the set of thermodynamic data corres- sensitivity with respect to our purposes of overall reliability
ponding to Model 3. The outputs recorded in Table 6 clearly and accuracy.
The Scientific World Journal 7

20 observe that HPLC-RP and HPLC-IEC are complementary

18 techniques: joining the separation ability reached with these
16 two methods we are now able to quantify up to 12 carboxylic
14 acids in wine. Conductometric results, together with the in-
Ethanol (%)

12
formation obtained by the synthetic “wines”, contributes to
10
validate our approach and assumptions. Switching from the
8
6
analytical to the equilibrium composition allows the pre-
4 diction of the effect on wines of oenological treatments (as
2 addition of acid-base active substances) or natural transfor-
0 mations (as precipitation or redox reactions). Many charac-
3 3.05 3.1 3.15 3.2 3.25 3.3 teristics of wine, specially those related to the ageing process,
pHcalc are always dependent from acid-base conditions. Moreover,
this investigation at molecular and thermodynamic level
Figure 2: Linear fitting applied in order to estimate the correct contributes to the basic knowledge, and the science presented
value of pH at the real value of % ethanol for Ch09. The equation
in this paper can also be used as an input for model spe-
model resulted y = 61.417x − 184.13 with R = 0.999.
ciation building of other natural fluids.
12.5
References
10.5 [1] C. S. Ough and M. A. Amerine, Methods for Analysis of Musts
and Wine, John Wiley & Sons, New York, NY, USA, 1988.
[2] E. Prenesti, S. Toso, P. G. Daniele, V. Zelano, and M. Ginepro,
8.5
“Acid-base chemistry of red wine: analytical multi-technique
pH

characterisation and equilibrium-based chemical modelling,”

6.5 Analytica Chimica Acta, vol. 507, no. 2, pp. 263–273, 2004.
[3] D. Tusseau and C. Benoit, “Routine high-performance liquid
chromatographic determination of carboxylic acids in wines
4.5 and champagne,” Journal of Chromatography A, vol. 395, no.
C, pp. 323–333, 1987.
2.5 [4] M. Castellari, A. Versari, U. Spinabelli, S. Galassi, and A.
0 0.5 1 1.5 2 2.5 3 3.5 Amati, “An improved HPLC method for the analysis of organic
V (mL) acids, carbohydrates, and alcohols in grape musts and wines,”
Journal of Liquid Chromatography and Related Technologies,
Figure 3: Alkalimetric titration curve (v = 25 mL) for natural () vol. 23, no. 13, pp. 2047–2056, 2000.
and synthetic () E08 at T = 25◦ C. Titrant: standardized KOH [5] D. Sanarico, S. Motta, L. Bertolini, and A. Antonelli, “HPLC
1.00 M. Natural E08: CH = 96.00 mM, pH = 3.05. Synthetic E08: determination of organic acids in traditional balsamic vinegar
CH = 95.76 mM, pH = 3.03. of Reggio Emilia,” Journal of Liquid Chromatography and
Related Technologies, vol. 26, no. 13, pp. 2177–2187, 2003.
To reach an experimental confirmation of our model, a [6] E. Prenesti, P. G. Daniele, S. Toso, V. Zelano, and S. Berto,
“Development of a speciation model for the interpretation
synthetic solution reproducing the composition of a wine,
of the acid-base properties of grape red wines,” Chemical
with respect to the acid-base reactivity (until pH < 7), was Speciation and Bioavailability, vol. 16, no. 1-2, pp. 17–24, 2004.
prepared. A mixture of carboxylic acids, proline, inorganic [7] P. G. Daniele, A. De Robertis, C. De Stefano, C. Rigano,
anions and cations, ethanol, and water was achieved (using and S. Sammartano, “Calcium acetate, succinate, maleate,
standardized solutions of each significant reactant) for E08, and phthalate complexes in aqueous solution, potentiometric
thus faithfully reproducing the content of bases (organic and determination of stability constants and their dependence on
inorganic ones) and the values of CH , ionic strength, per- ionic strength,” Annali di Chimica, vol. 73, pp. 619–633, 1983.
centage of ethanol. The synthetic mixture was alkalimetri- [8] “Analytical methods for wine analysis,” Journal of the European
cally titrated, as the wines. Figure 3 shows the overlap of the Communities, CEE n. 2676/90, annex 3, 1990.
[9] C. De Stefano, P. Princi, C. Rigano, and S. Sammartano,
titration curves for natural and synthetic E08 showing an
“Computer analysis of equilibrium data in solution. ESAB2M:
excellent agreement up to pH 7. an improved version of the ESAB program,” Annali di Chim-
ica, pp. 643–675, 1987.
4. Conclusions [10] C. De Stefano, P. Mineo, C. Rigano, and S. Sammartano,
“Ionic strength dependence of formation constants. XVII. The
The use of multitechnique analytical and equilibrium mea- calculation of equilibrium concentrations and formation con-
surements combined with a chemical modelling step allowed stants,” Annali di Chimica, vol. 83, pp. 243–277, 1993.
us to develop a thermodynamic approach to the acid-base [11] P. Fisicaro, E. Ferrara, E. Prenesti, and S. Berto, “Quality con-
chemistry of white wine. Since the chemical modelling is trol of pH measurements: uncertainty budget comparison
mainly based on the chemistry of carboxylic acids, particular of primary and secondary apparatuses,” in Combining and
attention was paid to quantify these substances. With respect Reporting Analytical Results, A. Fajgelj, M. Belli, and U.
to the past, we added HPLC-IEC to HPLC-RP to improve Sansone, Eds., pp. 96–103, Royal Society of Chemistry, Cam-
the characterization of the carboxylic fraction. We can bridge, UK, 2007.
 Photoenergy
International Journal of
International Journal of Organic Chemistry International Journal of Advances in
Medicinal Chemistry
Hindawi Publishing Corporation
International
Hindawi Publishing Corporation Hindawi Publishing Corporation
Analytical Chemistry
Hindawi Publishing Corporation
Physical Chemistry
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

International Journal of

Carbohydrate Journal of
Chemistry
Hindawi Publishing Corporation
Quantum Chemistry
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Submit your manuscripts at

http://www.hindawi.com

Journal of
The Scientific Analytical Methods
World Journal
Hindawi Publishing Corporation
in Chemistry
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of International Journal of International Journal of Journal of Bioinorganic Chemistry

Spectroscopy
Hindawi Publishing Corporation
Inorganic Chemistry
Hindawi Publishing Corporation
Electrochemistry
Hindawi Publishing Corporation
Applied Chemistry
Hindawi Publishing Corporation
and Applications
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014

Journal of  Chromatography   Journal of Journal of International Journal of

Theoretical Chemistry
Hindawi Publishing Corporation
Research International
Hindawi Publishing Corporation
Catalysts
Hindawi Publishing Corporation
Chemistry
Hindawi Publishing Corporation
Spectroscopy
Hindawi Publishing Corporation
http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014 http://www.hindawi.com Volume 2014