Analysis of polyphenols in white wine by CZE with amperometric detection using carbon nanotube-modified electrodes

Research Article

Analysis of polyphenols in white wine byCZE with amperometric detection usingcarbon nanotube-modified electrodes

A method for the simultaneous detection of five polyphenols (caffeic, chlorogenic, ferulic

and gallic acids and (1)-catechin) by CZE with electrochemical detection was developed.

Separation of these polyphenols was performed in a 100 mM borate buffer (pH 9.2)

within 15 min. Under optimized separation conditions, the performance of glassy carbon

(GC) electrodes modified with multiwalled carbon nanotube layer obtained from

different dispersions was examined. GC electrode modified with a dispersion of multi-

walled carbon nanotubes (CNT) in polyethylenimine has proven to be the most suitable

CNT-based electrode for its application as amperometric detector for the CZE separation

of the studied compounds. The excellent electrochemical properties of this electrode

allowed the detection of the selected polyphenols at 1200 mV and improved the effi-

ciency and the resolution of their CZE separation. Limits of detection below 3.1 mM were

obtained with linear ranges covering the 10�5 to 10�4 M range. The proposed method has

been successfully applied for the detection (ferulic, caffeic and gallic acids and (1)-

catechin) and the quantification (gallic acid and (1)-catechin) of polyphenols in two

different white wines without any preconcentration step. A remarkable signal stability

was observed on the electrode performance despite the presence of potential fouling

substances in wine.

Keywords:

Capillary electrophoresis / Carbon nanotubes / Electrochemical detection /Polyphenols / Wine DOI 10.1002/elps.201000498

1 Introduction

Polyphenols are strong antioxidant reagents that are

necessary for vegetal cell functioning and they can be found

in fruits and vegetables such as grapes, apples or onions as

well as in beverages as wine or tea. Nowadays, there is a

great expectation around these compounds as many studies

claim their beneficial effects on human health, such as

antioxidant, anti-thrombotic, anti-bacterial, anti-allergic and

anti-inflammatory activities [1]. As a consequence, the

intakes of polyphenol-rich foods are recommended to be

included in diet habits and new dietary supplements and

functional foods containing polyphenols have been released.

Since 1980, several studies have shown that lower risk of

chronic diseases was correlated with a diet rich in fruit and

vegetables [2, 3]. Therefore, an appropriate profiling of the

occurrence of such compounds in foods is required.

Wine is one of the foods with higher polyphenol

content. Its composition is very complex and most of the

compounds found come from grape and the fermentation

process. Polyphenols in grape are found in their skin,

especially in epidermis cells and seeds and their concen-

tration is significantly lower in pulp. Concentration and

variety of polyphenols in grape depends on many factors

such as vineyard type, wine type, climate conditions and

soil, season of harvesting, fermentation period in the

presence or absence of skin and seeds, maceration, ageing

etc. Polyphenols with antioxidant activity found in wine have

different structures and they can be derivatives from

phenolic acids, cynamic acids and tyrosine, stilbenoids or

flavonoids.

Therapeutic action of wine is related to its antioxidant

properties due to the presence of polyphenols since they

behave as scavengers of reactive oxygen species and metal-

chelators. Their beneficial activity after ingestion will

depend on their concentration, antioxidant power and

interaction with other components. Many studies correlate a

moderate wine intake (mostly red wine) with an increase in

Monica Moreno1

Alberto Sanchez Arribas1

Esperanza Bermejo1

Antonio Zapardiel2

Manuel Chicharro1

1Departamento de QuımicaAnalıtica y AnalisisInstrumental, UniversidadAutonoma de Madrid. Madrid,Spain

2Departamento de CienciasAnalıticas, Universidad Nacionalde Educacion a Distancia,Madrid, Spain

Received September 27, 2010Revised October 22, 2010Accepted October 22, 2010

Colour Online: See the article online to view Fig. 1 in colour.

Abbreviations: CNT, carbon nanotube; ED, electrochemicaldetection; GC, glassy carbon; GC/(CNT/PEI), GC electrodemodified with a dispersion of multiwalled CNT inpolyethylenimine; MWCNT, multiwalled carbon nanotube;

PEI, polyethylenimine

Correspondence: Dr. Manuel Chicharro, Departamento deQuımica Analıtica y Analisis Instrumental, Universidad Autono-ma de Madrid. C/ Francisco Tomas y Valiente, 7, 28049 Madrid,SpainE-mail: [email protected]: 134-91-497-4931

& 2011 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.electrophoresis-journal.com

Electrophoresis 2011, 32, 877–883 877

antioxidant power in plasma after polyphenol absorption

during the digestion process, which finally protects tissues

from oxidative stress.

Determination of the quantitative composition and

investigation of the factors affecting the composition of

bioactive substances, using reliable methods, is considered a

priority. Several methods have been developed to monitor

the level of phenolics in wine, such as the Folin–Ciocalteu

procedure and absorbance at 280 nm to produce a total

polyphenol measurement [4]; and HPLC for quantification

of individual monomeric [5, 6] and oligomeric [7, 8];

phenolics. However, these total phenol values do not indi-

cate the reducing strength of the polyphenols present, and

do not distinguish between catechol and galloyl-containing

polyphenols that are easier to oxidize (e.g. gallic and caffeic

acids, (1)-catechin, quercetin etc.) and polyphenols that are

more difficult to oxidize (e.g. coumaric acids and malvidin-

3-glucoside). Moreover, it is difficult to quantify both

monomeric and oligomeric phenolics in a single chroma-

tographic run.

HPLC has been the method of choice for the analysis of

phenolic compounds in wine, but CE has been increasingly

used for this purpose. The application of CE in the analysis

of beverages and foods [1, 3, 9, 10], including wine, and

specifically for resveratrol in wine [11], has been reviewed.

Comparison of the quantitative data obtained for phenolic

compounds in wine by HPLC and CE methods has been

carried out. While some small differences could be seen in

the results obtained for some phenolics by Garcıa-Viguera

and Bridle [12] and Wang and Huang [13], Andrade et al.[14] found no significant qualitative and quantitative differ-

ences in the results obtained by the two techniques. CE has

the advantages of high separation efficiency, short analysis

time, small sample size requirement, capability for minia-

turization and field application. The most used methodolo-

gies for the analysis of phenolic compounds in wine are

CZE and MEKC, although capillary isotachophoresis is also

used and, to a lesser extent, NACE [1]. Despite its many

advantages, electrochemical detection (ED) has been used

scarcely in CE separations for wine analysis [15–18], whereas

UV detection is the most widely used for this purpose.

In this work, we present the advantages provided by the

incorporation of carbon nanotube (CNT) films onto GC

electrodes used for ED of polyphenols in CZE separations of

wine samples. The advantages provided by CNTs when used

in ED systems coupled to CE have been recently reviewed

[19]. However, to our knowledge, this kind of application to

wine analysis using electrodes integrating these nanoma-

terials has not been reported to date. The closer relevant

applications have been reported by Chen and co-workers

who employed CNT composites for the CE analysis of

polyphenols in herbal extracts [20, 21] whereas Escarpa

and co-workers showed the utility of screen-printed

electrodes modified with CNT dispersions for the CE

analysis of polyphenols and water-soluble vitamins in apples

and pharmaceutical supplements using microchip platforms

[22, 23]. This CNT layer provides not only electrocatalytic

properties (typically resulting in diminution of overpotential)

but also an enhanced signal stability and increased resistance

to passivation, as it has been reported for analytes such as

NADH [24, 25] or phenol [26] that produce rapid surface

fouling in common electrodes when monitoring their elec-

trochemical oxidation during long periods of time. This

improvement in electrochemical signals is related to the

presence of a high density of edge-plane defects on the CNT

surface [24, 27], which enhances the electron-transfer kinetics

and impedes the apparition of fouling symptoms. In this way,

a remarkable improvement in stability and reproducibility of

signals during CZE analysis of wine samples is obtained and

high sample throughput is possible.

2 Materials and methods

2.1 Reagents and samples

Caffeic, chlorogenic, ferulic and gallic acids and

(1)-catechin were purchased from Sigma-Aldrich (Madrid,

Spain). The polyphenol stock solutions were prepared by

dissolving an appropriate amount of the compound in

methanol (Environmental grade, Alfa Aesar, Barcelona,

Spain). All the stock solutions were kept away from light

and stored under refrigeration. Diluted solutions were

prepared daily from these stock solutions. All the other

chemicals were analytical-reagent grade and they were

used without further purification. Ultrapure water

(r418 MO cm) from an Elga Purelab Option Q system

(ELGA LabWater, UK) was used for preparing all solutions.

Separation buffers were prepared from both the correspond-

ing 0.50 M phosphoric acid and boric acid solution and the

desired pH was set by adding 1.0 M NaOH and made up to

their final concentration. All the buffers and the samples

were submitted to ultrasonic treatment for 5 min and

filtered through 0.45 mm MFS-13 filters (Advantec MFS,

USA) before being introduced to the electrophoretic system.

White wine samples (Cumbre de Gredos and Jaume

Serra) were purchased in local markets and they were

analyzed within the first week after opening. Before being

analyzed, samples were filtered through ‘‘Osmonics’’

0.45 mm nylon filters (Micron Separations, MA, USA).

Multiwalled carbon nanotube (MWCNT) powder

(30715 nm diameter, 5–20 mm length, purity 495%,

‘‘hollow tube’’ type, prepared by chemical vapor deposition,

nominal 1% Fe and 0.1% S residuals, Lot. PD30L520-60805)

were obtained from NanoLab (MA, USA).

2.2 Apparatus

CE experiments were carried out with a P/ACE MDQ

(Beckman Coulter, Madrid, Spain) controlled by the soft-

ware 32 Karat 7.0. Amperometric detection was performed

with an Epsilon amperometric detector (BAS, West Lafay-

ette, IN, USA) connected to a Pentium IV PC computer with

Electrophoresis 2011, 32, 877–883878 M. Moreno et al.


ChomGraph 2.34.00 software (BAS) for data acquisition and

processing. No variations were introduced on the original

commercial setup of the CE equipment. Fused-silica

columns 55 cm-long (25 mm id and 365 mm od, Teknokro-

ma, Barcelona, Spain) were employed and no decoupling

strategy was followed. At the beginning of each day, the

capillary was conditioned by successive flushing with 1.0 M

NaOH, 0.10 M NaOH and separation buffer (5 min each).

Between runs, the capillary was rinsed consecutively with

0.10 M NaOH, water and separation buffer for 3 min.

Samples were introduced by hydrodynamic mode for 35 s

at 1.5 PSI (the injected volume being 6.6 nL). The electro-

chemical device designed for coupling to commercially

available CE equipments has been described previously [28]

but in this case the Plexiglas ground vial was not employed

and the capillary column was inserted directly inside the

electrochemical cell. The system could be mounted in a

short period of time (ca. 2 min), allowed an easy working

electrode replacement and the alignment and distance

between the working electrode and the capillary outlet could

be controlled without micropositioners. Silver and platinum

wires act as pseudo-reference and counter electrode,

respectively. The electrophoretic separations were carried

out at 251C while the detecting currents were allowed to

reach a stable baseline prior to amperometric monitoring.

pH was monitored with a 654 pH-meter from Metrohm

(Herisau, Switzerland).

2.3 Electrode preparation

GC electrodes of 1 mm diameter were from ALS (Cat no.

002412; Tokyo, Japan) and they were glued with non-

conductive Loctite 401 cyanoacrilate glue (Loctite, Spain) to a

conventional M6 plastic nut. Before using, they were

polished with 0.3 and 0.05 mm alumina slurries (Buehler;

Spain) on polishing cloths (Buehler) and subjected to

ultrasonic cleaning in water for 30 s. The electrode was

subsequently washed with pure water before using or

modifying GC modified with MWCNT dispersions were

prepared by dropping MWCNT dispersion (1 mL) onto clean

GC surface and left to dry for at least 2 h. Such MWCNT

dispersions were prepared by dispersing the appropriate

amount of as-received MWCNTs within 4.0 mL of disper-

sing solutions followed by ultrasonic treatment using

an ultrasonic probe Sonic Vibra Cell (model VCX750).

Three different dispersing media were tried: ethanol/water

(1:1 v/v), 0.5% w/v Nafion in ethanol/water (1:1 v/v) and

0.5% w/v polyethylenimine (PEI) in ethanol/water (1:1 v/v).

3 Results and discussion

3.1 CZE separation of polyphenols

Initially, CZE separation of the studied compounds was

evaluated using different separation electrolytes (phosphate,

borate and mixtures of them) at different pH (from 8 to 11)

in the absence and in the presence of an organic

modifier (methanol) in different proportions (5, 10, 20 and

30%). Amperometric detection was employed for these

studies using a bare GC as working electrode and a

detection potential of 1600 mV. In accordance with pKa of

the studied compounds, they are negatively charged under

these conditions and their separation is achieved by the

different deprotonation rate of their phenolic moieties and

the formation of complexes between catechol groups and

borate ions [29]. The most favorable separation conditions

were 100 mM borate buffer of pH 9.4 at 25 KV, as a

compromise between resolution, efficiency and analysis

time (15 min). The presence of phosphate and methanol did

not enhance resolution significantly and analysis time

became unsuitably long.

The amount of sample injected was optimized in order

to achieve a favorable signal-to-noise ratio (S/N) without

losing the resolution. It is well known that electrokinetic

sample introduction induces a selective enrichment of some

sample constituents depending on their charge; therefore,

the hydrodynamic mode was preferred in order to keep

sample composition. The best analytical performance was

obtained applying 1.5 PSI for 35 s.

3.2 Amperometric detection using MWCNT-modi-

fied electrodes

The performance of GC electrodes modified with a MWCNT

layer from different dispersions was examined by compar-

ing the hydrodynamic voltammograms (Fig. 1) of the

studied polyphenols registered under the CZE separation

conditions selected in Section 3.1. At bare GC, only ferulic

acid and (1)-catechin shown signals for detection potentials

lower than 1400 mV and just when it was higher than

1600 mV it was possible to obtain a clear signal of all

compounds. In addition, no obvious plateau was attained

before 1900 mV except for ferulic acid. This situation

substantially changed when an MWCNT layer was incorpo-

rated onto the GC surface. In all cases, all the compounds

showed a well-defined signal at 1400 mV and a plateau was

reached before 1600 mV. A closer inspection of these

results reveals that the charge of the dispersing polymer

played an important role in the potential shift observed

when GC was modified using each type of MWCNT

dispersion. MWCNT layers obtained from positively charged

PEI dispersions showed a more pronounced diminution

in oxidation potentials with hydrodynamic waves of

(1)-catechin and gallic, chlorogenic and caffeic acids

starting in the 0 to 1100 mV interval and reaching a

plateau at 1200 mV. On the contrary, the MWCNT layer

from Nafion (negatively charged polymer) dispersion

provided a moderate decrease in the oxidation potentials

for (1)-catechin, gallic acid and especially for caffeic and

chlorogenic acids. In the absence of dispersing polymer

(ethanol dispersion), an intermediate potential shifting was

Electrophoresis 2011, 32, 877–883 CE and CEC 879


observed. These results could be partially explained since

analytes are negatively charged under these separation

conditions so they could suffer electrostatic repulsion or

attraction with Nafion or PEI, respectively. These electro-

static interactions can be the reason why lower over-

potentials associated with notably higher amperometric

signals were observed with MWCNT layers obtained from

PEI dispersions. Another important feature is the impress-

ive decrease in oxidation potentials (about 600 mV) of

analytes containing catechol-like groups, namely, caffeic,

chlorogenic and gallic acids and (1)-catechin, in the

presence of MWCNT when compared to bare GC. The

electrochemical oxidation of these groups appear to be

catalyzed when using MWCNT, as previously suggested by

Pumera and co-workers [30], while ferulic acid, which

contains only an hydroxyl substituent in its aromatic ring,

displayed a moderate decrease in its oxidation potential

(about 200 mV) when compared to bare GC. In our case, the

combination of electrostatic interactions induced by the

dispersing polymer and the electrocatalytic properties of

MWCNT can lead to a favored orientation of analytes

resulting in such decrease in overpotentials.

In light of the results obtained, GC/(CNT/PEI) has

proven to be the most suitable CNT-based electrode for its

application as amperometric detector for the CZE separation

of the studied compounds. Figure 2 shows the electro-

pherograms obtained for the caffeic, chlorogenic, ferulic

and gallic acids and (1)-catechin, in their CZE separation

under optimal separation conditions, using a bare GC and

a GC/(CNT/PEI) electrode and applying 1200 mV as

detection potential. Well-defined peaks and a very good base-

line separation were achieved for the five polyphenol

compounds when a GC/(CNT/PEI) was employed, whereas

no peak was obtained in a bare GC electrode when using

this detection potential. Improvement in the efficiency and

resolution observed when the separation was monitored

with GC/(CNT/PEI) should be noted. Indeed, the peak

widths are notably smaller when compared to those

observed at a bare GC working at 1600 mV (Fig. 2B). This

effect can be a consequence of the superior electrochemical

efficiency of CNT that oxidize the analytes before they can

diffuse towards the electrode edge. Therefore, the obtained

peaks should be narrower than those provided by the less-

effective bare GC [31]. Increase in the sensitivity of the

signals obtained with GC/(CNT/PEI) is another feature

related to the good electrochemical qualities of CNT.

3.3 Method performance

The analytical characteristics of the amperometric detection

(AD) of polyphenols using GC/(CNT/PEI) were evaluated using

a series of standard mixture solutions of the five analytes with

concentrations ranging from 1.0� 10�5 to 2.0� 10�4 mol/L.

The results obtained are summarized in Table 1. There is an

excellent correlation between the peak current and the

concentration of each analyte within the concentration range

indicated in Table 1. The limits of detection (LODs) for the five

analytes studied, evaluated on the basis of an S/N of 3, ranged

from 2.3 to 3.2 mM. (0.45–1.1 mg/L).

The stability and reproducibility of the amperometric

signals were evaluated by performing eight successive CZE

runs of a standard mixture solution (2.0� 10�5 mol/L for

each analyte) in different days (n 5 4), replacing the elec-

trode surface each day. Acceptable RSD values (r14%) of

peak currents were obtained in all cases (Table 1) in this

inter-electrode comparison, reflecting the appropriate

reproducibility in the CNT-casting process. Intra-electrode

repeatability was estimated using the RSD values (n 5 8)

calculated each day and they varied from 2.3 to 4.8% for (1)-

catechin whereas values from 3.5 to 7.1% were calculated for

-100 100 300 500 700 900 -100 100 300 500 700 900

GC

5 nA

GC/EtOH/CNT

GC/Nafion/CNT

GC/PEI/CNT

Potential (mV)

Figure 1. Hydrodynamic voltammograms of caffeic (square),chlorogenic (circle), ferulic (up triangle) and gallic (downtriangle) acids and (1)-catechin (diamond) at different electro-des. Separation buffer, 100 mM borate buffer of pH 9.4.Separation voltage, 25.0 KV. Hydrodynamic sample introduction,1.5 PSI, 35 s. Polyphenols concentration, 100 mM.

0.0 2.0 4.0 6.0 8.0 10.0 12.0 14.0

45

32

b

Time (min)

a

8.0 nA

1

A

B

Figure 2. CZE separations of (1)-catechin (1) and chlorogenic(2), ferulic (3), caffeic (4) and gallic (5) acids under optimalseparation conditons: (A) using a bare GC electrode andapplying a detection potential of 1600 mV; (B) using a bare GC(a) and GC/(CNT/PEI) (b) electrodes and applying a detectionpotential of +200 mV. Polyphenols concentration, 100mM.



ferulic acid. No apparent signal decay was noticed during

these experiments. Therefore, CNT-modified electrode

performance was not affected by surface fouling. The same

experiment was carried out for comparison at bare GC

electrode using 1600 mV as detection potential and an

important signal decrease was observed after each run. The

consequence of this fouling process was the suppression of

82% of the initial signal after eight runs. This remarkable

improvement in signal stability conferred by the presence of

the CNT layer supports the reliability of long-term analysis

of polyphenol samples without electrode conditioning, thus

enhancing the reproducibility as well as saving time.

These results demonstrated that the GC/(CNT/PEI)

electrode is suitable for the detection of the five analytes,

allowing a reproducible and more sensitive quantification of

them in samples containing these compounds in the

concentration levels described previously.

3.4 Determination of polyphenols in white wine

samples

In order to demonstrate the practical application of the GC/

(CNT/PEI) electrode, the determination of polyphenols in

different samples of white wine was conducted by CE-AD

under the selected optimum conditions. Wine is considered

as a rich source of phenolic acid and flavonoids but the

phenolic amount varies considerably in the different wine

types, depending on the grape variety, environmental factors

in the vineyard and the wine processing techniques. Several

examples can be found in the literature in which the

polyphenols studied have been determined in samples of

white wine at levels above the LODs of the methodology

developed in this work [32–34]. Thus, the possibility of direct

analysis of untreated wine sample was evaluated. Figures 3a

and 4a show the typical electropherograms obtained for the

two white wine varieties analyzed. Several peaks can be

observed in both cases; thereby the identification of

polyphenol compounds was necessary. Comparison of the

migration time of these peaks with those obtained in the

electropherogram of the standard mixture solution as well

as the application of the standard addition approach were

employed for this purpose. In this way (1)-catechin was

identified in the Cumbre de Gredos sample whereas ferulic,

caffeic and gallic acids were in the Jaume Serra sample.

With the aim of estimating the concentration of the

polyphenols in these wine samples, several additions of the

Table 1. Performance characteristics of the CE-AD method developed

Analyte Linear range (M) Regression equation LOD (mM)a) RSD (%)b)

(1)-Catechin 1.0� 10�5�1.4� 10�4 I(nA) 5 (0.270.2)1(7.470.6)� 104 C(M), r 5 0.9992 3.1 9

Chlorogenic acid 1.1� 10�5�1.4� 10�4 I(nA) 5 (�0.170.3)1(7.370.6)� 104 C(M), r 5 0.990 3.2 11

Ferulic acid 7.6� 10�6�8.0� 10�5 I(nA) 5 (�0.270.7)1(1.070.2)� 105 C(M), r 5 0.990 2.3 14

Caffeic acid 9.2� 10�6�1.6� 10�4 I(nA) 5 (0.570.3)1(8.170.5)� 104 C(M), r 5 0.993 2.8 12

Gallic acid 8.2� 10�6�1.6� 10�4 I(nA) 5 (�0.270.4)1(9.170.6)� 104 C(M), r 5 0.990 2.5 10

a) The detection limits correspond to concentrations giving a signal-to-noise of 3.

b) RSD (%) value of current signal, based on the mean signal of eight different injections of standard mixture (2.0� 10�5 M each) in

4 days (n 5 4).

0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0

c

b

Time (min)

5.0 nA

a

1

Figure 3. Electropherograms of standard additions of (1)-catechin (peak 1) to Cumbre de Gredos wine sample. Additionsof (1)-catechin, 0 (a), 21 (b) and 43 mM (c). Working electrode,GC/(CNT/PEI). Detection potential, 1200 mV. Separation condi-tions as in Fig. 1.

0,0 2,0 4,0 6,0 8,0 10,0 12,0 14,0 16,0 18,0 20,0

d

c

b

54

3

2

Time (min)

5,0 nA 1

a

Figure 4. Electropherograms of standard additions of polyphe-nols to Jaume Serra wine sample. Additions of chlorogenic,ferulic, gallic and caffeic acids and (1)-catechin, 0 (a), 10 (b), 20(c) and 40 mM (d) each. Working electrode, GC/(CNT/PEI).Detection potential, 1200 mV. Separation conditions as in Fig.1. Peak identification as in Fig. 2.



standard solution of polyphenols were performed (Figs. 3

and 4). The standard addition plots were built and the

results obtained are summarized in Table 2. In all cases the

slopes calculated from these plots were similar to those

obtained in the external calibration (Table 1). In view of the

results obtained, the quantification of (1)-catechin

(1171 mg/L, n 5 5) and gallic acid (2.270.2 mg/L, n 5 5)

was possible in Cumbre de Gredos and Jaume Serra,

respectively. These results were in accordance with values

reported (from 0.4 to 25 mg/L) elsewhere [32–34] for the

analysis of these polyphenols in other varieties of white

wines. Another remarkable fact was the excellent stability

observed on the electrode performance despite the presence

of potential fouling substances in wine. The signals

obtained for a CZE run of a standard mixture solution

(2.0� 10�5 mol/L for each analyte) performed before and

after the wine sample analysis did not show significant

differences and the decrease in peak signals was lower than

9%. These results indicate that this method provides a

useful quantitative method for the direct analysis of poly-

phenols in white wines.

4 Concluding remarks

This work demonstrates that GC/(CNT/PEI) showed

excellent electrochemical performance for the detection of

(1)-catechin, caffeic, gallic, chlorogenic and ferulic acids in

their CZE separation. These results proved that this GC/

(CNT/PEI) can be a valuable tool for amperometric

detection of polyphenols in CE. The studied analytes can

be detected below 3.5 mM when using the proposed

methodology and their direct detection in white wine

samples was possible. The excellent electroanalytical

properties of these GC/(CNT/PEI) anticipate that useful

methodologies for the polyphenol mapping and character-

ization of wine types can be readily developed in connection

to CE.

The authors wish to thank the Ministerio de Ciencia eInnovacion (Grant CTQ 2009-09791) for the financial support.

The authors have declared no conflict of interest.

5 References

[1] Hurtado-Fernandez, E., Gomez-Romero, M., Carrasco-Pancorbo, A., Fernandez-Gutierrez, A., J. Pharm.Biomed. Anal. 2010, 53, 1130–1160.

[2] Kris-Etherton, P. M., Hecker, K. D., Bonanome, A., Coval,A. E., Binkoski, A. E., Hilpert, A. E., Griel, A. E., Etherton,T. D., Am. J. Med. 2002, 113, 71–88.

[3] Cifuentes, A., Electrophoresis, 2006, 27, 283–303.

[4] Escarpa, A., Gonzalez, M. C., Anal. Chim. Acta 2001, 427,119–127.

[5] Lamuela-Raventos, R. M., Waterhouse, A. L., Am. J.Enol. Vitic. 1994, 45, 1–5.

[6] Price, S. F., Breen, P. J., Valladao, M., Watson, B. T., Am.J. Enol. Vitic. 1995 46, 187–194.

[7] Lazarus, S. A., Adamson, G. E., Hammerstone, J. F.,Schmitz, H. H., J. Agric. Food Chem. 1999, 47,3693–3701.

[8] Kennedy, J. A., Waterhouse, A. L., J. Chromatogr. A2000, 866, 25–34.

[9] Sadecka, J., Polonsky, J., J. Chromatogr. A 2000, 880,243–279.

[10] Garcia-Canas, V., Cifuentes, A., Electrophoresis 2008,29, 294–309.

[11] Gu, X., Chu, A., O’Dwyer, M., Zeece, M., J. Chromatogr.A 2000, 881, 471–481.

[12] Garcıa-Viguera, C., Bridle, P., Food Chem. 1995, 54,349–352.

[13] Wang, S. P., Huang, K. J., J. Chromatogr. A 2004, 1032,273–279.

[14] Andrade, P. B., Oliveira, B. M., Seabra, R. M., Ferreira,M. A., Ferreres, F., Garcıa-Viguera, C., Electrophoresis2001, 22, 1568–1572.

[15] Du, F., Fung, Y. S., Electrophoresis 2010, 31, 2192–2199.

[16] Peng, Y., Chu, Q., Liu, F., Ye, J., J. Agric. Food. Chem.2004, 52, 153–156.

[17] Scampicchio, M., Wang, J., Mannino, S., PrakashChatrathi, M., J. Chromatogr. A 2004, 1049, 189–194.

[18] Yu-Hua, C., Zang, S., Yu-Zhi, F., Jian-Jong, Y., Chem. J.Chin. U. (Gaodeng Xuexiao Huaxue Xuebao) 2001, 22,2011–2013.

[19] Pumera, M., Escarpa, A., Electrophoresis 2009, 30,3315–3323.

[20] Wei, B., Wang, J., Chen, Z., Chen, G., Chem. Eur. J.2008, 14, 9779–9785.

Table 2. Determination of polyphenols in white wine samples

White

wine

Analyte Slope

(L/mol nA)a)

(n 5 5)

Foundb)

(mM) �(mg/L)

RSD

(%)c)

Cumbre (1)-Catechin (6.670.5)� 104 (3874) � (1171) 11

de

Gredos

Chlorogenic

acid

nd nd nd

Ferulic acid nd nd nd

Caffeic acid nd nd nd

Gallic acid nd nd nd

Jaume (1)-Catechin nd nd nd

Serra Chlorogenic

acid

nd nd nd

Ferulic acid (1.170.1)� 105 Detected 18

Caffeic acid (7.170.6)� 104 Detected 16

Gallic acid (9.270.5)� 104 (1271)–(2.070.2) 12

a) Slope evaluated for five different wine samples.

b) Mean and standard deviations of five determinations.

c) RSD (%) value of peak current signal, based on five different

injections of wine samples.

nd, not detected.



[21] Chen, Z., Zhang, L., Chen, G., Electrophoresis 2009, 30,3419–3426.

[22] Gonzalez Crevillen, A., Avila, M., Pumera, M., Gonzalez,M. C., Escarpa, A., Anal. Chem. 2007, 79, 7408–7415.

[23] Crevillen, A. G., Pumera, M., Gonzalez, M. C., Escarpa,A., Lab Chip 2009, 9, 346–353.

[24] Banks, C. E., Compton, R. G., Analyst 2005, 130,1232–1239.

[25] Kachoosangi, K. T., Musameh, M. M., Abu-Yousef, I.,Yousef, J. M., Kanan, S. M., Xiao, L., Davies, S. G.,Russell, A., Compton, R. G., Anal. Chem. 2009, 81,435–442.

[26] Arribas, A. S., Bermejo, E., Chicharro, M., Zapardiel, A.,Luque, G. L., Ferreyra, N. F., Rivas, G. A., Anal. Chim.Acta 2007, 596, 183–194.

[27] Ambrosi, A., Sasaki, T., Pumera, M., Chem. Asian J.2010, 5, 266–271.

[28] Sanchez Arribas, A., Moreno, M., Bermejo, E., Lorenzo,M. A., Zapardiel, A., Chicharro, M., Electrophoresis2009, 30, 3480–3488.

[29] Kuhn, R., Hoffstetter-Kuhn, S., Capillary Electrophor-esis: Principles and Practice, Springer-Verlag, Heidel-berg 1993, pp. 93–94.

[30] Crevillen, A. G., Pumera, M., Gonzalez, M. C., Escarpa,A., Analyst 2009,134, 657–662.

[31] Wang, J., Tian, B., Prakash Chatrathi, M., Escarpa, A.,Pumera, M., Electrophoresis 2009, 30, 3334–3338.

[32] Woraratphoka, J., Intarapichet, K.-O., Indrapichate, K.,Food Chem. 2007, 104, 1485–1490.

[33] Minussi, R., Rossi, M., Bologna, L., Cordi, L., Rotilio, D.,Pastore, G. M., Duran, N., Food Chem. 2003, 82,409–416.

[34] Peres, R. G., Micke, G. A., Tavares, M. F. M., Rodriguez-Amaya, D. B., J. Sep. Sci. 2009, 32, 3822–3828.



Documents

Analysis of polyphenols in white wine by CZE with amperometric detection using carbon nanotube-modified electrodes