Vitis 43 (4), 205–208 (2004)

Titrimetric method based on potentiometric titration to evaluate

redox couples in wine and polyphenols

N. VIVAS , M.F. NONIER and N. VIVAS DE GAULEJAC

Demptos Cooperage Posted to CESAMO (Centre d’Etudes Structurales et d’Analyses des Molécules Organiques),
Université Bordeaux I, Talence, France
.

Summary When analyzing wine, rather than quantifying total

polyphenols (SANTOS-BUELGA and WILLIAMSON 2003, VIVAS
Polyphenols are electroactive compounds, each mol- et al. 2003) it is more appropriate to evaluate each redox
ecule having reductive (Rd) and oxidative forms (Ox) due to polyphenol group in order to follow the evolution of the
the presence of several phenolic hydroxyls. Direct study of wine during oxidation and to estimate its oxido-reductive
Rd/Ox forms was conducted using potentiometric titration capacity by a global index (VASCONCELOS et al. 1999,
in two steps: First, a complete reduction of the polyphenol KILMARTIN et al. 2001). In research laboratories the oxidative
Rd/Ox couples by titanium III chloride TiCl3 (100 % Rd), status of molecules under different conditions (solution
and then, oxidative titration by dichlorophenolindophenol composition, pH, reagent concentration, metallic catalyst)
DCPIP (100 % Ox). The second curve represents the total- is monitored. In enology, potentiometric titration of oxido-
ity of polyphenol couples titration with, for each couple, a reductive compounds in wine was demonstrated by
characteristic E0 point representing the equilibrium be- RIBEREAU-GAYON and GARDRAT (1957). These authors showed
tween the concentration of Rd and Ox forms (Rd/Ox = 1). that it is possible to quantify and measure the different re-
This value was typical for each polyphenol. We conducted dox polyphenols and their respective normal E0 potential.
several preliminary experiments with a model polyphenol Potentiometric titration is a two-step process. First, the
(catechin) to establish the analysis conditions. The titra- polyphenol or wine solution is completely reduced by a se-
tion solution concentration was N/10 in HCl N and N/20 in lected reducer. Thereafter, all redox couples are in a reduced
water for TiCl3 and DCPIP respectively. The concentration form. Subsequently, the same solution is oxidized by a se-
of this solution diminished very rapidly, i.e. within 24 h for lected oxidizing agent. A second curve is obtained that dis-
TiCl3 and DCPIP (under nitrogen and in darkness) and plays all the redox compounds of a sample with the expres-
regularly needed a fresh preparation. To control the con- sion of each characteristic E0 value.
sistency and precision of the method we performed a few
tests on pure products (e.g. hydroquinone/quihydrone, Fe
III, Fe II, Cu II) permitting comparison between theoretical Material and Methods
and experimental E0 values. Our result indicated less than
5 % variation and curves with a high reproducibility. All solutions were prepared with ultrapure water from a
Millipore Milli-Q system. HCl purum quality (Merck), alco-
K e y w o r d s : Wines, polyphenols, oxido-reduction, hol (absolute, Merck Eurolab), L(+)-tartaric acid (>99 % pu-
potentiometric titration, dichlorophenolindophenol, EO. rity, Acros Organics), NaOH (98 %, SDS) were used as
hydroalcoholic solution for sample dilution before analysis
(adjustment of pH at the initial pH of wine). The composi-
Introduction tion of the hydroalcoholic solution was 12 % vol. EtOH,
5 g·l-1 tartaric acid and NaOH to adjust the pH. (+)-catechin
Polyphenols are electroactive compounds which are and (-)-epicatechin hydrates (Sigma Aldrich, 93.7 % purity,
involved in many different oxidation or reduction reactions by H1 NMR, Bruker DPX300) were used as wine polyphenol
affecting the composition and quality of food and bever- models in order to determine the method’s parameters (1 g·l-1
ages (WEENY et al. 1974, TANIZAWA et al. 1984, MAYER 1987, in hydroalcoholic solution at pH 3.5). The phenols were two
LATTANZIO et al. 1989, HIROSE et al. 1990). This property is isomeric forms of flavan-3-ols. TiCl3 (reducer agent, 10 % in
due to phenolic hydroxyls that depend on the polymeric HCl 20-30 %, Sigma Aldrich) was prepared in 1N HCl (Sigma
level of the molecules. Oxidation by enzymatic or chemical Aldrich) and DCPIP (oxidizing agent, dichlorophenolindo-
reactions induces polyphenol polymerization and intensi- phenol sodium hydrate, Sigma Aldrich) in ultrapure water.
fies browning (WATERS 1964). On the basis of these enzyme The titrator was a DL50 (Mettler Toledo) with two inter-
oxidation properties, many investigations have contributed changeable 10 ml burettes (DV910), a potentiometric elec-
to the determination of polyphenols by amperometric trode (DM140-SC) combined with a platinum ring electrode
biosensors (PRAVDA et al. 1995, EGGINS et al. 1997). for redox titration in a range of 0-70 °C, with 3M KCl satu-

Correspondence to: Dr. N. VIVAS, Demptos Cooperage Posted to CESAMO (Centre d’Etudes Structurales et d’Analyses des Molécules
Organiques), Université Bordeaux I, 351, Cours de la Libération, 33405 Talence, France. Fax : +33-5-4000-26 23. E-mail:
n.vivas@cesamo.u-bordeaux1.fr
206 N. VIVAS et al.

rated with AgCl reference electrolyte. The electrode was T i t r a t i o n a g e n t s l i f e t i m e : Titration

calibrated with a redox buffer solution and washed over- agents in solution lost their reductive or oxidative power
night in concentrated ammonia and a few min in an ultra- when stored. To determine the lifetime of agents, we titrated
sonic bath with water. The whole system was controlled by a (+)-catechin hydroalcoholic solution (1 g·l-1) over time.
computer. Preparation of samples: 50 ml of hydroalcoholic The results collected in Fig. 1 show that for TiCl3 and DCPIP
solution of (+)-catechin, other polyphenols or 50 ml of wine the correct lifetime (i.e. without any change in the titration
diluted at 1/50 in the same solution for reduction. Titration curves) was 24 h; the best preservation conditions were in a
conditions were: ml per 20 s for a reduction increment of 0.03 dark room at ambient temperature. After 24 h, some crystals
and ml per 30 s for an oxidation increment of 0.02. Titration were formed in the aqueous DCPIP solution. After 10 d, all
started only after nitrogen saturation of the samples and all the solutions were completely de-titrated.
experiments were managed under a nitrogen atmosphere.
Temperature: 20 °C ±1 °C.

1.00
Results and Discussion

Volume of titration agents, ml

C h o i c e o f t i t r a t i o n a g e n t s : A reducer
0.75
and an oxidant were used for the potentiometric titration.
The selected products did not affect the polyphenol struc-
tures and allowed a back-titration with no changes com-
pared to the first titration. Several agents were tested: so- 0.50
dium hydrosulfite for the reduction was unstable in solu-
tion; for oxidation, hydrogen peroxide was very unstable;
potassium permanganate is a stronger oxidant destroying 0.25
polyphenols, and iodine causes the formation of strong com-
plexes with polyphenols. Finally, we retained two agents:
titanium III chloride (TiCl3) as reducer and dichlorophenol-
indophenol (DCPIP) as oxidant. 0 12 24 48 72
For TiCl3, solutions prepared at N/100 to N were tested Time of conservation, h
during a titration of a (+)-catechin solution (1 g·l-1). At N/100, Fig. 1: Titration agents as function of time of conservation at room
N/80 reductive power was too low, at N/50, N/30 the reduc- temperature in the dark. White circles corresponding to TiCl3,
tion was incomplete and slow (> 1 h), at N/10 the reduction black circles to DCPIP. Mean values of three replicates.
went to completion in 15 min and consequently was se-
lected for the experiment. In accordance with the following Repeatibility of the method: A
equation of reaction, 1 mole of TiCl3 reduced one phenolic repetition of (+)-catechin titration yielded satisfactory re-
ketone in phenols: sults (Tab. 1). Under our conditions, for a pure product, the
maximum variability of the curves was < 5 % when replica-
O OH
tions were made in succession; however, when replicate ti-
R OH + H+ + e- R OH tration was performed several days later, variability was
higher, but no more than 8 %. Statistical results were similar
Ti+++ Ti++++ + e-
for wine: < 7 % variability for successive replications and
O OH < 10 % after some time.
+ +++ + Ti++++
R OH + H + Ti R OH
Table 1

For DCPIP, in order to respect the stoichiometry of the Precision of titration assay for (+)-catechin solutions. EHmin,
global reaction, the N/20 solution was retained. With these minimim of EH at the end of reduction; EHmax, maximum of EH at
conditions, titration of (+)-catechin solution (1 g·l-1) was the end of oxidation. Vmax, total volume of titration solutions for
a complete titration
performed in 25 min. In accordance with the following equa-
tion of reaction, 1 mole of DCPIP oxidized 2 phenolic OH
group: Reductiona Oxidationb
EHmin Vmax EHmax Vmax
OH O (mV) (ml) (mV) (ml)

R OH R O + 2H+ + 2e-
Average (n = 6) -328 0.825 348 1.02
DCPIP + 2H+ + 2e - DCPIPH2 Standard deviation 5.01 0.04 2.38 0.08
Confidence interval 4.01 0.03 1.9 0.06
OH O (for a 5 %)
R OH + DCPIP R O + DCPIPH2
a TiCl3, b DCPIP.
 Titrimetric method to evaluate redox couples 207

I m p o r t a n c e o f p H v a l u e : Before analysis, C o m p a r i s o n o f t h e o r e t i c a l a n d e x-
the samples need to be diluted in order to shorten analysis p e r i m e n t a l E 0 v a l u e s : The normal E0 potential
time and to limit the volume of titration agents required. To was typically that of constant redox couples, some of which
reproduce a composition similar to wine, we chose an have been reported in literature (ATKINS 1990). E0 was calcu-
hydroalcoholic medium. The pH values of this preparation lated in accordance with the NERNST law (VIVAS et al. 1996).
significantly affected the curve profiles obtained with (+)-cat- Fig. 3 shows experimental E0 determination of (+)-catechin
echin (Fig. 2). Particularly, with increasing pH we noted a and (-)-epicatechin, which were different (275 and 171 mV,
faster reduction, mainly at pH 4.5. For average wine values respectively). It is interesting to note that the epicatechin,
(3.5-4.0), there was little variation. The difference was greater with the lower E0 value, is more oxidized than catechin (FREITAS
for oxidation than for reduction, even in the pH range of et al. 1996). In Tab. 2, we show three examples: two mineral
wine. Normal E0 potential was not affected. To standardize redox couples and one organic couple, a simple phenol
the conditions of the assay, polyphenols were all analyzed quinhydrone/hydroquinone indicating high agreement be-
at pH 3.5. For wines, the dilution medium was prepared at tween theoretical and measured E0 values.
the same pH as the titrated wine.

400 400 3.5

4.0
3.0 2.5

4.5
200 200
EH, mV

EH, mV

0 0

-200 -200
3.0
2.5
3.5
4.0
4.5
-400 -400
1 2 V, ml 1 2 V, ml

Fig. 2: Influence of pH on the titration curves in reduction (left graph) and oxidation (right graph) of (+)-catechin in hydroalcoholic media.
pH values ranged from 2.5 to 4.5. V (ml) is the volume of titration agent (TiCl3 N/10 for reduction, DCPIP N/20 for oxidation), EH (mV)
is the oxidoreduction potential.

a 2 b
350

2 400
150
1
EH, mV

∆EH, mV

300
1
-50
200

-250
100

-450 0
0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0
V, ml V, ml

Fig. 3: Potentiometric titration of (+)-catechin (1) and (-)-epicatechin (2). a: Characteristic curve (white symbols: reduction, black
symbols: oxidation). b: Titrogramme obtained by calculation of ∆EH. E0 of (+)-catechin and (-)-epicatechin were 275 and 171 mV,
respectively.
208 N. VIVAS et al.

Table 2 References

Values of normal potentials E0 obtained by theoretical calcula- ATKINS , P. W.; 1990: Physical Chemistry. Oxford University Press,
tion and experimental measurements Oxford.
EGGINS, B.; HICKEY, C.; TOFT, S. A.; ZHOU, D. M.; 1997: Determination
of flavanols in beers with tissue biosensors. Analyt. Chem. Acta
Eo Variation 347, 281-288.
FREITAS, V.; GLORIES, Y.; LAGUERRE, M.; 1996: Oxydation des procyanidines
Theoretical Experimental %
des pépins dans un milieu modèle et dans le vin. In: A. LONVAUD-
FUNEL (Ed.): Oenologie 95, 375-380. Lavoisier, Tec&Doc, Paris.
Fe++/Fe+++a 770 710 ± 15 7.8 HIROSE, Y.; YAMAOKA, H.; NAKAYAMA, M.; 1990: Oxidation product of
Cu+/Cu++a 150 160 ± 5 6.7 epicatechin under radical reaction. Yukagaku 39, 967-969.
KILMARTIN, P. A.; ZOU, H.; WATERHOUSE, A. L; 2001: A cyclic voltammetry
Quinhydrone/ method suitable for characterizing antioxidant properties of wine
hydroquinoneb 699.5 713 ± 24 1.9 and wine phenolics. J. Agric. Food Chem. 49, 1957-1965.
LATTANZIO, V.; LINSALATA, V.; PALMIERI, S.; VAN SUMERE, C. F.; 1989: The
a titration in distilled water. beneficial effect of citric and ascorbic acid on the phenolic brown-
b titration in hydroalcoholic solution at pH 3.5. ing reaction in stored artichoke (Cyanara scolymus L.) heads.
Food Chem. 33, 93-106.
MAYER, A. M.; 1987: Polyphenols oxidases in plants. Recent progress.
A p p l i c a t i o n t o a l c o h o l i c b e v e r- Phytochemistry 26, 11-20.
a g e s a n d v i n e g a r : The method was applied to PRAVDA, M.; JUNGAR, C. M.; IWUOHA, E. I.; SMYTH, M. R.; VYTRAS, K.;
different samples of wines, vinegar and brandies; satisfac- IVASKA, A.; 1995: Evaluation of amperometric glucose biosensors
based on co-immobilisation of glucose oxidase with an osmium
tory results were obtained in all cases (Fig. 4). The method redox polymer in electrochemically generated polyphenols films.
was also tested on different polyphenol sources and pure Analyt. Chim. Acta, 304, 127-138.
molecules; it proved to have a general validity. R IBEREAU -G AYON , J.; G ARDRAT , J.; 1957: Application du titrage
potentiométrique à l’étude du vin. Ann. Technol. 2, 185-216.

350
300

200
150

50
EH, mV
EH, mV

-100
-150

-300 -250

-450
0.2 0.4 0.6 0.8 1.0 1.2 0.2 0.4 0.6 0.8 1.0 1.2
V, ml V, ml

Red wine Port Wine (Vintage) Sweet White wine (Sauterne)

Red wine vinegar Brandy (Armagnac)

Fig. 4: Potentiometric titration curves for wines, brandy and red wine vinegar. On the left: reduction, on the right: oxidation. pH of dilution
solutions was adjusted to the pH of samples.

SANTOS-BUELGA, C.; WILLIAMSON, G.; 2003: Methods in Polyphenol Analy- VIVAS, N. ; VIVAS DE GAULEJAC, N. ; NONIER, M. F. ; 2003: Sur l’estimation
sis. Royal Soc. Chem., Cambridge. et la quantification des composés phénoliques des vins. Bull.
TANIZAWA, H.; TODA, S.; SAZUKA, Y.; TANIYAMA, T.; HAYASHI, T.; ARICHI, S.; O.I.V., 865-866, 281-303.
TAKINO, Y.; 1984: Natural antioxidants. I. Antioxidative compo- W ATERS , W. A.; 1964: Mechanisms of Oxidation of Organic Com-
nents of tea leaf (Thea sinensis L.). Chem. Pharm. Bull. 32, pounds. John Wiley & Sons Inc, New York.
2011-2014. WEENY MC, D. J.; KNOWLES, M. E.; HEARNE, J. F.; 1974: The chemistry
VASCONCELOS, M. T.; AZENHA, M.; FREITAS V.; 1999: Role of polyphenols of non-enzymatic browning in foods and its control by sulphites.
in copper complexation in red wines. J. Agric. Food Chem., 47, J. Sci. Food Agric. 25, 735-746.
2791-2796.
VIVAS, N.; GLORIES, Y.; BERTRAND, A.; ZAMORA , F.; 1996: Principe et
méthode de mesure du potentiel d’oxydoréduction dans les vins. Received April 15, 2004
Bull. O.I.V., 785-786, 617-633.