J. Agric. Food Chem.

2006, 54, 607−616 607

Determination of Total Content of Phenolic Compounds and

Their Antioxidant Activity in VegetablessEvaluation of
Spectrophotometric Methods
PAVEL STRATIL, BOŘIVOJ KLEJDUS, AND VLASTIMIL KUBAÄ Ň*
Department of Chemistry and Biochemistry, Mendel University of Agriculture and Forestry,
Zemedelská 1, CZ-613 00 Brno, Czech Republic

This research studies in detail the contents of phenolic compounds determined by the Folin-Ciocalteu
reagent and the antioxidant activities determined by the TEAC (Trolox equivalent antioxidant capacity),
DPPH (using diphenyl-p-picrylhydrazyl radical), and FRAP (ferric reducing antioxidant power) methods,
and their correlations for used standards with these methods (catechine, gallic acid, caffeic acid,
ferulic acid, Trolox, ascorbic acid, and ferrous sulfate) and extracts from several species of commonly
consumed vegetables were studied in detail. The comparison of absolute values of absorption
coefficients for used standards and for individual methods allows one to choose optimal common
standards for methods to be compared. The procedures applied for the same sets of the extracts
using identical calibration procedures and common standards allowed better comparison of the results
obtained by the TEAC, DPPH, and FRAP methods. The values of content of phenolic substances
and total antioxidant activity of the sets of samples correlate very well for all used methods. The very
high values of antioxidant activity were found in intensely colored vegetables (red cabbage, red onion,
etc.), and the values were very low in watery vegetables such as potato, marrow, and cucumber.

KEYWORDS: Vegetables; phenolic compounds; antioxidant activity; spectrophotometry; unified stan-

dardization

INTRODUCTION in many cases, a basic role in their pharmacological effects;

Phenolic antioxidants, a specific group of secondary metabo- thus, it can be considered the most important (9). The basic
lites, play the very important role of protecting organisms against plant constituents of food should be a sufficient source of
harmful effects of oxygen radicals and other highly reactive phenolic compounds for humans. However, there is a limited
oxygen species. Their formation in human organisms is closely amount of information on the content of phenolic compounds
connected with the development of a wide range of degenerative in common foodstuffs of plant origin and their antioxidant
and nondegenerative diseases, mainly arteriosclerosis and other activities.
associated complications, cancer, indispositions, and last but not The content of phenolic compounds in plant materials can
least with the accelerated aging of organisms (1-4). be determined by several separation methods [high-performance
Some preventive and defensive systems against the attack of liquid chromatography (HPLC), gas chromatography (GC),
the reactive substances exist in the human organism; however, capillary zone electrophoresis (CZE), etc.] as a set of individual
they cannot eliminate harmful activities of such substances substances or by a specific chemical reaction as a group of
completely, particularly when their production is increased in chemically similar reactive compounds. The chromatographic
some metabolic, physiologic, pathologic, and other situations. determination is highly precise and accurate with high informa-
An adequate intake of natural antioxidants in food is therefore tive value, but it is highly problematic to identify all of the
of great importance for protection of macromolecules against phenolic compounds in a specific plant material at acceptable
oxidative damage (4, 5) in cells (mainly unsaturated fatty acids times and costs of analysis. The determination of antioxidant
in lipids, cholesterol, different functional polypeptides and activity by HPLC is experimentally more complicated, and at
proteins, and nucleic acids). present, procedures are mostly presented to show methodological
Plants are rich in phenolic compounds of different origins
principles.
and functions. Most of them belong to the principal, biologically
highly active components (antiviral, anticancerogenic, etc.) of Generally, it is necessary to analyze several samples since
foodstuffs of plant origin (2, 6-8). Antioxidant activity plays, the contents of phenolic substances can vary in different varieties
and cultivars. In addition, contents of phenolic substances are
* To whom correspondence should be addressed. Tel: +420-545 133 influenced by a large number of external factors such as
285. Fax: +420-545 212 044. E-mail: kuban@mendelu.cz. agrotechnical processes, climatic conditions and ripeness during
10.1021/jf052334j CCC: $33.50 © 2006 American Chemical Society
Published on Web 01/11/2006
608 J. Agric. Food Chem., Vol. 54, No. 3, 2006 Stratil et al.

harvest, postharvest manipulations, and time of consummation Table 1. Comparison of Our Contents of Phenolic Compounds
(10). At the present time, sufficient information for the evalu- Determined by FC Method and Dry Matter (DM) with Published
ation of content of total phenolic compounds and total antioxi- Results of Kähkönen et al. (31) and Vinson et al. (16)
dant activity as a criterion of nutritional value of plant foodstuffs
content of phenolic compounds
can be easily obtained from spectrophotometric measurements
employing specific analytical reagents. totala freeb totalb DM
The determination of antioxidant activity of plant extracts is x x ± sx x x %
still an unresolved problem. Approximately 20 analytical n vegetable Sc 31 Sc 16 Sc 16 Sc 16
methods such as applying different reagents, reaction mixture 1 carrots (garden) 9.9 0.6 ± 0.0 5.4 4.7 20.3 15.3 19.8
composition, standards, analytical evaluations, and others are 2 carrots (Delvita 14.5 13.9 29.9 12.5 10.6
used. The exact comparison of the results and their general stores)
interpretation are practically impossible due to the variability 3 parsley 10.9 6.3 22.5 36.1
of experimental conditions and differences in physicochemical 4 celery 10.6 7.6 7.0 21.7 13.6 13.1 8.8
5 red beet 20.7 4.3 ± 0.2 20.7 45.2 42.6 53.4 22.0 15.4
properties of oxidizable substrates. Furthermore, the antioxidant 6 red radish 15.0 0.9 ± 0.1 28.7 30.9 3.9
activity of the substances in foodstuffs and other biological 7 onion yellow 18.2 2.5 ± 0.1 12.4 4.7 37.5 22.9 16.2 10.5
systems depends on the applied test system (method) and 8 onion red 21.2 3.0 ± 0.9 25.5 11.4 43.6 41.0 13.6 9.8
substrate that should be protected by the antioxidative substance. 9 garlic 18.9 10.5 5.2 38.9 34.3 35.2 37.6
10 leek 13.7 11.3 28.3 13.2
A huge number of factors, including colloidal properties of 11 potato, var. 4.5 4.3 ± 0.2 5.8 3.5 9.3 5.9 24.6
substrate, experimental conditions, reaction medium, oxidation Laura (red)
state, and antioxidant localization in different phases, can 12 potato, var. Korela 6.8 2.5 ± 0.1 8.1 13.9 19.1 20.3
influence the activity. The applied methods can be divided into 13 potato, var. Katka 5.4 7.4 11.2 23.9
two basic groups: (i) the test with lipophilic substrates and (ii) 14 lettuce 26.8 20.3 8.4 55.2 16.9 3.4 4.7
15 spinach paste 15.5 20.4 13.4 32.0 27.6 7.2 6.2
the test with hydrophilic substrates. Furthermore, the composi- 16 cabbage white 12.1 10.5 9.5 25.0 19.2 11.6 9.4
tion of the system, type of oxidizable substrate, method used, 17 cabbage red 36.3 129.0 74.7 13.6
and the way of quantification of antioxidant activity are of great 18 cabbage green 13.5 11.1 27.9 18.9
importance during the tests (11). 19 cabbage Chinese 11.0 5.3 22.6 5.5
20 cauliflower 14.0 10.8 11.7 28.7 20.9 11.2 8.6
Each determination should be done at different experimental 21 kohlrabi green 10.0 7.9 20.5 13.3
conditions of the oxidation reaction using at least two or more 22 broccoli 18.1 18.2 17.5 37.3 40.6 10.1 8.9
methods for quantification of different products of oxidative 23 tomato 10.7 2.0 ± 0.1 9.4 9.5 22.0 18.9 8.3 6.9
reaction, and finally, the general trends of values for individual 24 pepper green 15.1 18.4 16.4 31.2 26.1 8.1 5.7
25 pepper red 19.3 23.1 39.7 11.1
samples should be compared. In addition, a unified standardiza- 26 pepper red 16.5 25.6 34.0 11.0
tion of the antioxidant activity test is highly recommended (11). (Spain)
Hydrolyzed extracts contain more active antioxidants since it 27 pepper red 12.6 32.6 25.9 12.5
is well-known that the glycosides of phenolic substances are (Spain)
weaker antioxidants than the corresponding aglycones (12, 13). 28 cucumber 10.6 3.8 ± 0.1 7.7 4.4 21.9 15.5 5.0 2.6
29 marrow 7.7 6.6 5.6 15.9 17.6 5.6 5.1
The main purpose of the present study was to determine the 30 little bean, 10.6 1.6 ± 0.1 3.9 21.8 25.6
total content of phenolic compounds in selected species of frozen
commonly consumed vegetables and to determine and compare 31 little bean green, 10.5 6.2 8.0 21.6 17.8 11.0 8.4
the antioxidant activity of plant extracts applying three com- frozen
monly used spectrophotometric methods: TEAC (Trolox equiva- 32 maize delicates, 12.7 7.8 13.9 26.2 19.1 33.5 25.7
lent antioxidant capacity), FRAP (ferric reducing/antioxidant frozen
power), and DPPH (with 2,2-diphenyl-1-picrylhydrazyl assay).
a GE, gallic acid equivalent in mg/g DM. b CE, catechin equivalent in µmol/g
The analytical procedures were modified to allow semimicros-
DM. c Our results.
cale measurements at the same experimental conditions using
a large set of extracts with acceptable reproducibility. The
selection of a suitable, generally applicable standard for all of tocopherol, purum, g99%, for HPLC), Folin-Ciocalteu reagent (FC
methods allowed us to obtain a set of simply comparable results. reagent), and 2,4,6-tris(2-pyridyl)-s-triazine (TPTZ, puriss, g99.0%)
were from Fluka Chemie (Buchs, Switzerland). Methanol and aceto-
nitrile of gradient grade were from Merck (Darmstadt, Germany). Other
MATERIALS AND METHODS chemicals of p.a. purity were from Pliva-Lachema (Brno, Czech
Instruments. A spectrophotometer HELIOS β (declared reliability Republic). All reagents and standard solutions were prepared using Milli
of measuring to 2.0 AU) controlled with program VISION 32 Software Q deionized water (Millipore, Bedford, MA).
(Spectronic Unicam, Cambridge, United Kingdom), an ultrasonic bath Sample Preparation. Samples (26 kinds of vegetables commonly
HD 2070 (model M8 72, Bandelion Sonoplus, Germany), a high-speed consumed in the Czech Republic; see Table 1) were purchased at a
mill Grindomix (Retsch, Germany), and a CHRIST ALPHA 1-2 B local market (Delvita stores). Edible parts of the vegetables (20-50 g)
lyophiliser (Braun Biotech International, Germany) were used for were cut into small peaces (<0.5 cm) and lyophilized at -52 °C for
sample preparation. An HP 1100 chromatographic system (Hewlett- 48-72 h in a vacuum. The dry matter was determined gravimetrically
Packard, Waldbronn, Germany) with a mass-selective HP MSD detector ((1 mg) by the difference before and after lyophilization. Lyophilizates
(G1946A, Hewlett-Packard, Palo Alto, CA) was used for ascorbic acid were homogenized in a laboratory ultramixer to tiny particles and sieved
determination. through a plastic sieve (<0.5 mm). The powder was stored in plastic
Chemicals. Caffeic acid, ferulic acid, ascorbic acid (purity g 99.0% bottles under nitrogen at -20 °C until analyses. Lyophilization of the
each), 2,2-diphenyl-1-picrylhydrazyl radical (DPPH•, ≈90.0%), and native sample was assumed to be the most cautious method for isolation
2,2′-azino-bis-(3-ethylbenzothiazolin-6-sulfonate) diammonium salts of phenolic compounds, and it was preferred as the less cautious
(ABTS, ≈98.0%) were purchased from Sigma-Aldrich Chemical Co. homogenization of fresh samples (14). It minimized aerial and
(United States); gallic acid monohydrate (g98.0%), Trolox (6-hydroxy- enzymatic oxidation of phenolic compounds that are more intense at a
2,5,7,8-tetramethylchromane-2-carboxylic acid, a hydrophilic derivative higher content of water and at laboratory temperature.
Phenolic Content and Antioxidant Activity in Vegetables J. Agric. Food Chem., Vol. 54, No. 3, 2006 609

Extraction and Hydrolysis. A modification of procedure for or plant extracts (from 5 to 25 µL according to reaction intensity) were
extraction of phenolic compounds published by Vinson et al. (15, 16) mixed with the working solution (975 µL) and diluted up to 1000 µL
was employed. Briefly, the homogenized powder (2 × 500 mg; (0.2 with deionized water. A decrease of absorbance was measured at 734
mg) was weighed and equally divided into two plastic bottles with screw nm after 20 min. Aqueous phosphate buffer solution (1 mL, without
cups. An aqueous methanol (1:1, v/v, 10 mL) was added into one bottle ABTS•+ solution) and Trolox (1.0-2.0 mmol/L) were used as a control
to extract free phenolic compounds while the same amount of acidified and main calibrating standard, respectively.
aqueous methanol (1:1, v/v, 2.4 mol/L HCl, 10 mL) was added into FRAP Method. A FRAP method (25-27) was modified to a
the second bottle to extract all phenolic compounds (free and conjugated semimicroscale with a total volume of 1 mL. Briefly, a portion of an
usually with saccharide moiety). The suspensions were sonicated under aqueous 10 mM solution of TPTZ reagent in 40 mmol/L HCl was mixed
nitrogen atmosphere for 2 min at 90% power for better destruction of with the same volume of 20 mmol/L FeCl3‚6H2O and 10 times higher
cell walls and then incubated at 82-83 °C (bp of the extraction agents) volume of acetate buffer of pH 3.6 (3.1 g sodium acetate and 16 mL
for 150 min with vortexing every 30 min for 20 s. After incubation, acetic acid per liter). The mixture was incubated at 37 °C for several
methanol (10 mL) was added into the cooled samples, the mixture was minutes. A portion (900 µL) of the Fe3+-TPTZ mixture and 20-50
vigorously agitated, and the suspension was centrifuged at 6000g for µL of the sample (or standard or water for blank) were diluted up to
10 min. Acidic extracts were neutralized to pH 7 since results of some 1000 µL with deionized water and incubated for at least 4 min (better
methods were pH-dependent. Supernatants were stored at -20 °C under 10-30 min) (27), and the absorbance was measured at 593 nm. Fresh
nitrogen in plastic vials with screw caps. Extracts were stable for at 1 mmol/L working solutions of FeSO4 were used for calibration
least several days or several weeks at 4 and -20 °C, respectively. (dilution of 20 mmol/L stock solution). The antioxidant capacity was
Determination of Phenolic Compounds in Plant Extracts. The calculated from the linear calibration curves (25, 28). The total
Folin-Ciocalteu method (FCM), based on the reduction of a phos- antioxidant activity further increased seriously (26, 27) between 4 and
phowolframate-phosphomolybdate complex by phenolics to blue 30 min (more then 40-50% in most cases).
reaction products, was used to determine phenolic compounds (16- DPPH• Test/Method. An original procedure (29) was modified (to
18). The total volume of reaction mixture was miniaturized to 1 mL. 1 mL total volume) with calibration using standard solution. The
An absorbance was measured twice (four or six times in some cases) working solution was prepared by dilution of methanolic DPPH• (98
for each sample at 760 nm against blank (100 µL water) using 2 mmol/L mg/L, absorbance = 1.9) to absorbance at = 1.5 AU (exactly the same
gallic and ferulic acids and (+)-catechine as standards. Five points for sample and standard) to gain the sufficient reaction capacity for
calibration was linear to a concentration of 0.2 mmol/L in the reaction higher contents of antioxidants in extracts. A portion (950 µL) of the
mixture and an absorbance range up to 3.0 AU. Highly reproducible working solution and 5-50 µL of the sample in water or (1:1, v/v)
results for standards (R2 > 0.997) and samples were obtained except methanolic extract of phenolics (or standard) were diluted up to 1000
for some samples with high concentrations of phenolics, partly due to µL with aqueous methanol (1:1, v/v) and incubated for at least 20 min
dilution and dosing of small volumes (5 µL) of extracts. (better 30 min to reach plateau) (30), and the absorbance was measured
Ascorbic acid and main mono- and disaccharides (glucose, fructose, at 515 nm against methanol and methanol with a DPPH• blank.
and saccharose) present in vegetables and fruits can interfere. Corre-
sponding corrections for ascorbic acid interferences are experimentally RESULTS AND DISCUSSION
complicated due to its instability and due to fast subsequent reactions.
In addition, most of the published methods for ascorbic acid determi- As can be seen from the Figure 1A-D, the rate of the
nation are unsuitable for plant extracts. Thus, the HPLC/MS method reactions of individual standards and samples differs substan-
was applied. Interferences of saccharides were less evident than those tially, and the reaction kinetics of samples can be expressed by
of ascorbic acid (see parameters of calibration curves), and their exponential curves reaching the plateau after 30-60 min, where
influence for the samples with higher contents of saccharides could be about 95% yield is obtained. This reaction time corresponds to
estimated on the basis of the mean contents of phenolics and saccharides the optimum incubation period. The best time for absorbance
(18).
measurements was near 30 min for all four methods. Further
Determination of Antioxidant Activity/Capacity of Plant Ex-
tracts. Three methods, TEAC, FRAP, and DPPH•, based on reaction
prolongation of the reaction time led to some other disadvantages
with electron-donating or hydrogen radicals (H•) producing compounds/ in some cases, i.e., adsorption of color products on the test tube
antioxidants according to the reaction R• + Aox-H f RH + Aox•, or cuvette walls or evaporation of methanol and thus the increase
were used. Electron transfer and hydrogen atom transfer reactions can of absorbance values (concentration of productssthis phenom-
be difficult to distinguish. Hydrogen atom transfer reactions can be enon can be prevented by using stoppered cuvettes).
the result of proton-coupled electron transfer. Despite the similar redox Choice of a Suitable Common Standard for the Tested
mechanisms of the methods, reagents and products are different. Also, Methods. Strictly linear calibration curves were obtained for
performance and interpretation differ considerably in publications. Thus, all of the methods using different calibration standards (gallic,
reciprocal comparison of results of individual methods and results ferulic, and caffeic acids, catechine, Trolox, and ferrous sulfate)
among publications for the same method is often problematic.
and main interferents (ascorbic acid, glucose, fructose, and
Trolox and ferrous sulfate are usually used for calibration of the
TEAC and FRAP methods, respectively, while the DPPH• method is
saccharose) in concentrations up to 50, 100, or 200 µmol/L of
usually interpreted on the basis of antioxidant amount needed for the individual compounds in dependence of their antioxidant
decrease of the initial DPPH• concentration to 50% (EC50); thus, they activity, up to 100 mmol/L of saccharides, and up to absorbance
are the initial DPPH• concentration dependent. A more objective values of 2.0-3.0 AU as long as the reagent was not completely
comparison of the results could be possible by applying the same depleted. The linearity of calibration curves allowed quantifica-
interpretation procedure with the same common standard and unified tion of phenolics using any of the above-mentioned standards.
standardization procedure. Different calibration standards and different ways of expres-
TEAC Method. The TEAC (or total antioxidant activity) method sion of concentrations (dry weight, DW; fresh weight, FW; in
(19-22) was modified. Briefly, the total volume of the original molar or mass units) have been applied to express the results
analytical procedure (23, 24) was reduced to 1 mL. A (1:1, v/v) mixture in the literature. These facts complicate the comparison of the
of ABTS (7 mmol/L) and potassium persulfate (4.95 mmol/L) was left
results from one source to another one.
to stand for 12 h at laboratory temperature in the dark to form radical
cation ABTS•+. The solution was stable for at least 1 week at 4 °C in For example, the content of phenolics (free and total) has
the dark. been expressed mainly as gallic (GE) or ferulic acids (FE) or
A working solution was diluted to absorbance values between 1.0 catechine (CE) equivalents in molar units or using a mean value
and 1.5 AU at 734 nm with phosphate buffer solution (constant initial of molecular weight (of phenolics, 290.0; of catechine, 298.3;
absorbance values must be used for standard and samples). Standards and of gallic acid, 170.1) using DW or FW of analyzed samples.
610 J. Agric. Food Chem., Vol. 54, No. 3, 2006 Stratil et al.

Figure 1. Rate of the reactions of the DPPH (A), FCM (B), FRAP (C), and TEAC (D) methods. A: 0, DPPH alone; 1, extract of plumb (free phenolics);
2, extract of peach (25 µL); 3, extract of peach (50 µL); and 4, extract of plumb (total phenolics). B: 1, ferulic acid; 2, extract of orange; and 3, gallic
acid (two injections). C: 1, extract of peach; 2, extract of red pepper; 3, extract of green pepper; and 4, reagent alone. D: 1, reagent alone; 2, extract
of green beans; and 3, extract of carrot.

Figure 2. Comparison of the absolute values of mmol absorption coefficients for used standards, phenol, and ascorbic acid determined by the FCM,
TEAC, FRAP, and DPPH methods.

A recalculation of determined values in different publications of (+)-catechine react with FC reagent while one reacting -OH
and their comparison is possible if the values of DW (or FW, group is free in the case of ferulic acid (18). The final absorbance
usually given), parameters of calibration equations (mostly not values are usually proportional to the number of reacting
given), and type of calibration standards are presented. The dry phenolic hydroxyl groups and also depend on a structure of
matter values of the samples of vegetables varied in the range the molecule, i.e., reactivity of individual phenolic hydroxyl
of 4-36% in our experiments and were in very good agree- (Figure 2).
ment with those in the literature (16), with few exceptions For example, two of three reacting hydroxyl groups of gallic
(Table 1). acid are less reactive in the FC method then two reacting
Some interpretations of values determined by individual hydroxyl groups of caffeic acid. The absorbance value for caffeic
method can be deduced from the comparison of the absolute acid (two reacting OH) is approximately twice and for (+)-
values of molar absorptivities of samples (Figure 2 and Table catechine (three reacting OH) three times higher than that for
2). Two of three hydroxyl groups of gallic acid and three of phenol (one reacting OH). A higher value was found for ferulic
four free hydroxyl groups (five -OH groups are present in total) acid (one reacting OH) than for phenol (substituents of the
Phenolic Content and Antioxidant Activity in Vegetables J. Agric. Food Chem., Vol. 54, No. 3, 2006 611

Table 2. Parameters of Calibration Equations A ) a × c + b for the explained by the fact that the content of phenolic compounds
FC, FRAP, TEAC, and DPPH Methods in some species of vegetables can be influenced by several
internal and external factors. The contents of conjugated
coefficients of calibration equationa A ) a × c + b phenolics ranged from 23 to 87% for most of the vegetables.
compound FCMb FRAPb TEAC DPPH Significantly higher (lower) values of free phenolics were found
gallic acid 0.0177 0.1005 −0.1588 + 1.4875 −0.1398 + 1.4511 for seven (four) types of vegetables while very similar results
caffeic acid 0.0201 0.0715 −0.0179 + 1.4668 −0.0279 + 1.3195 were obtained for another seven species from the 18 analyzed.
ferulic acid 0.0145 0.0471 −0.0624 + 1.0204 −0.0244 + 1.4363 Slightly or significantly higher contents of total phenolics were
catechin 0.0298 N N N
Trolox 0.0058 0.0396 −0.0235 + 1.0921 −0.0217 + 0.9278
found for 13 samples, and lower contents were found for five
Troloxc N 0.0388 −0.0261 + 1.4325 −0.0254 + 1.4215 samples.
phenol 0.0101 NR −0.0174 + 1.3440 NR The content of free and total phenolic compounds for 30
FeSO4 0.0028 0.0198 −0.0099 + 1.3771 −0.0158 + 1.4763 samples (two out-layers) correlates highly and significantly (r
ascorbic acid 0.0128 0.0282 −0.0172 + 0.9906 −0.0147 + 1.3080
glucose 0.0006 NR −0.0002 + 1.4266 NR ) 0.8092). The published values from different sources can
fructose 0.0022 0.0004 −0.0006 + 1.3563 NR seriously differ as can be seen from the values for our red beet,
saccharose 0.0005 NR −0.0011 + 1.4848 NR which was measured to be five times higher than that of
Kähkönen (31), but our value is still about 20% lower than that
a A, absorbance; c, concentration (µmol/L, for saccharides in mmol/L); N, not
of Vinson et al. (16). The total content of phenolic compounds
repeated; NR, no reaction. b b ) 0. c Separate experiment. in three varieties of pepper ranged between 810 and 1430 CE
mg/kg (CE ) catechine equivalent) and was similar to the data
benzene ring decrease the electron density on the hydroxyl of Karakaya (32). Colored varieties of vegetables (red onion,
group). The value is lower for gallic acid (two reacting OH red cabbage, and red pepper) are especially rich in phenolic
groups) than the doubled value for phenol (the liberation of the compounds. Red onion contains a higher quantity of both
second electron/hydrogen radical is more difficult). colored and colorless phenolics than the yellow variety. This is
The absolute values of molar absorption coefficients of FRAP in correspondence with the findings of HPLC analyses of the
methods were proportionally increasing with the number of main flavonole of onion, quercetine (33). In addition, they also
reacting phenolic hydroxyl groups, and their reactivity cor- contain anthocyanins, the strongly absorbing phenolic colorants
responded to the influence of the molecular structure on the that represent their red color.
electron density of the particular phenolic group. A significant The contents of phenolic compounds decrease in the order
decrease of the overall reactivity of phenolic hydroxyls of caffeic red cabbage, garlic, red beet, delicates maize, parsley, red onion,
acid was observed for the TEAC method. The reactivity of the yellow onion, green pea, white cabbage, pepper, and broccoli
hydroxyl group of Trolox and ferulic acid was nearly the same while the lowest values were found for marrow, cucumber,
for TEAC and DPPH methods, and the reactivity of the second Chinese cabbage, and radish (Figure 3).
hydroxyl group of caffeic acid was reduced while the reactivity
Determination of Antioxidant Activity/Capacity of Ex-
of all three hydroxyl groups of gallic acid was increased (Figure
tracts. Determination by TEAC Method. The TEAC method is
2 and Table 2).
one of the most often used methods for the determination of
Measured values by particular methods are therefore deter-
total antioxidant capacity (19-22). It is based on a neutralization
mined by selected standards. If the standard used for calibration
of radical cation formed by a single-electron oxidation of a
is highly reactive and gives a high absorbance, then the
synthetic ABTS chromophore to a strongly absorbing ABTS•+
measured values of samples will be low. A standard with
radical (700-750 nm) according to the reaction ABTS - e-
different reactivity by particular method will give very different
f ABTS•+. The radical is prepared by an oxidation reaction of
values for sets of samples measured by a particular method.
ABTS with oxidized methmyoglobine by enzymatic peroxidase
According to the results in Figure 2, Trolox seems to be the
reaction with H2O2 (original method) or (11) with potassium
best standard for all three methods, since nearly the same results
persulfate (higher sensitivity).
were obtained for the TEAC and DPPH methods and about 50%
higher values were obtained for the FRAP method. Slightly less The radical reacts quickly with electron/hydrogen donors to
suitable reaction properties were observed for the other stan- form colorless ABTS. The reaction is pH-independent. A
dards. On the other hand, for the FCM method, caffeic acid decrease of the ABTS•+ concentration is linearly dependent on
(eventually gallic acid or catechine) seems to be the very suitable the antioxidant concentration, including Trolox as a calibrating
standard while Trolox is not suitable due to relatively low standard (23, 34, 35). The results correlated very well with
reactivity and low values of absorbance values. biological redox properties of phenolics (36, 37).
Determination of Phenolic Compounds in Plant Extracts. The highest, however, the less reproducible values of anti-
The contents of phenolic compounds (free and total) determined oxidant activity were obtained in comparison with the other
by the FC method for different analyzed vegetables are presented tested methods. The ABTS• reagent was very unstable, and it
in Figure 3 as the corrected values for ascorbic acid interfer- was slowly degraded at the given experimental conditions. The
ences. Conjugated phenolic compounds prevailed in the veg- measurable continuous decrease of absorbance was observed
etables, and their contents varied between 33.7 and 82.2% with during the incubation period (c. 30% per hour; Figure 1D).
the exception of one sort of pepper (24.9%) and radish (7.1%). While the reaction with Trolox (standard) was very fast (reaction
Contents higher (lower) than 50% were found for 20 (10) species time about 1 min) and was constant for at least 30 min (not
of vegetables, respectively, with a maximum value for little presented in figures), the reactions with plant extracts were slow
beans at 82.2%. The extreme value of 7.1% for radish might with exponential form of the time dependence. The relatively
be seriously influenced by the presence of sulfur-containing very long time was needed to reach a plateau (even more than
compounds and may be an artifact. 1 h in most cases) although most of the reaction product was
Our results are mostly similar to those for similar sets of formed in the first 10 min.
vegetables and fruits (Table 1) published by Vinson et al. (16). The continuous decrease of sample absorbance could be
Higher deviations among the results in some cases can be affected by spontaneous degradation of ABTS• reagent (they
 612
J. Agric. Food Chem., Vol. 54, No. 3, 2006

Figure 3. Content of total and free phenolic compounds determined by the FC method (GE) and antioxidant activity determined by the TEAC, FRAP, and DPPH methods (TE) in extracts of vegetables in mmol/kg
fresh mass (the numbers correspond to those in Table 1). x ± absolute mean deviation, n ) 2, *n ) 4, **n ) 6), and TEAC × 2 for no. 9.
Stratil et al.
Phenolic Content and Antioxidant Activity in Vegetables J. Agric. Food Chem., Vol. 54, No. 3, 2006 613

closely correlate). The different rate of the reactions of ABTS• The determined values in extracts of vegetables were usually
with Trolox and plant extracts could not eliminate the decrease slightly higher than those obtained by DPPH method; however,
in absorbance values due to degradation of ABTS• reagent. A they were several times lower than those of the TEAC method.
steady linear decrease of absorbance after 7 min in the presence They correlated highly significantly with them (Figure 3) with
of some individual compounds (i.e., ascorbic acid) or an the exception of red cabbage.
exponential decrease for some other compound (i.e., quercetin) Determination by DPPH• Method. The DPPH• test/method
was published (34). A measured spontaneous exponential is one of the oldest and the most frequently used methods for
decrease of ABTS• reagent absorbance, an exponential decrease total antioxidant potential/capacity of food extracts (41, 42). It
of absorbance at reaction of ABTS• with extracts (Figure 1D), is based on the ability of antioxidant to give hydrogen radical
and the decrease in absorbance (0.538 AU) for 20 µmol/L to synthetic long-lived nitrogen radical compounds DPPH•
Trolox are in agreement with the published data (24, 38). having a radical localized on the N-atom: DPPH• + Aox f
From the above-mentioned facts, it could be deduced that DPPH, i.e., dN - N•- + H• f ) N - NH-.
the determined values for samples should be corrected (de- A blue-violet color changes gradually to green and yellow
creased) by the values of absorbance decrease obtained for the (absorption maximum at 405 nm), and a decrease in absorbance
pure ABTS• reagent. The values without this correction are at 515 nm is monitored during the reaction in neutral medium.
overestimated. The reaction is pH-dependent. Published results are based on
The antioxidant activity values determined by TEAC method the very different interpretation; therefore, comparison is very
for individual species of vegetables decreased in the order red problematic both in rank of the method and among others
beet > garlic > freezed maize delicates > red cabbage > parsley methods. Two versions of DPPH• method of evaluation (39)
> white cabbage > red onion > green pea > potato > yellow have been used, (i) dynamic and (ii) static.
onion > leeks > broccoli > spinach > pepper, etc., while the The rate of DPPH• destruction after addition of a sample
lowest values in increasing order are for cucumber < marrow containing phenolic compounds is measured in dynamic version.
< kidney bean < cabbage < lettuce < tomato < Chinese The reaction rate characterizes the reactivity, and it can be
cabbage, etc. (Figure 3). calculated as a slope of the reaction kinetics at the very
Determination by FRAP Method. A FRAP method is based beginning of the reaction (t ∼ 0). Usually, applied graphical
on the ability of antioxidant to reduce (electron transfer) Fe3+ interpretations are of limited accuracy and precision. In addition,
to Fe2+ ions in the presence of TPTZ forming an intense blue the starting reaction rate and the total antioxidant activity are
Fe2+-TPTZ complex with an absorption maximum at 593 nm. not completely two independent values.
The reaction is pH-dependent (optimum pH 3.6). The absor- The amount of DPPH• inactivated by the sample is measured
bance decrease is proportional to the antioxidant (reductant) graphically or numerically [as a % of consumed or unconsumed
content (25). Compounds that are active in Fe3+ reduction also DPPH•, % of discoloration 100 × (1 - Asample/Acontrol) etc.] in
stimulate the formation of OH• (prooxidation activity). a static version. The H-donor potential, in the form IC50 (29),
An absorbance exponentially increased due to the formation expressed percentage decrease of initial concentration (EC50),
of absorbing product of the reaction of antioxidants with Fe2+- i.e., amount of antioxidant needed for decrease of DPPH•
TPTZ complex. The reaction with extracts was slow, and the concentration by 50%. Direct comparison of the results in rank
constant and quantitative yield of the reaction plateau was not of the method and among other methods is again highly
reached even after 1 h (Figure 1C). The results of our problematic (39) since they depend on the initial concentration
observation are in agreement with some published data (27). of DPPH• radical.
Ferrous sulfate is usually used as a standard; however, the Radical scavenging efficiency (RSE, the ratio of initial
application of more widely used Trolox (TEAC and DPPH) can reaction rate of DPPH• radical decomposition to EC50) was
be advantageous due to linear dependence of absorbance vs recommended for characterization of radical inhibition ability
antioxidant concentration and its, approximately, twice higher (43). Antiradical efficiency (AE), expressed in two forms, AE
absorbance values as compared to FeSO4. The reaction with ) 1000 × 1/IC50 or AE ) 1/IC50‚t50, where t50 is the time
Trolox reached the maximum in 1 or 2 min, and the absorbance needed for 50% transformation of DPPH• radical, were also
is constant up to 30 min. recommended for presentation of results (29, 44, 45). It is clearly
The results of the FRAP method correlated very well with evident that the parameters are hardly reproducible, since the
the antioxidant activity measured by electron spin resonance values depend on the initial DPPH• concentration.
spectroscopy (ESR), i.e., the ability to donate electron or The percentage of DPPH• inhibition (I %), I % ) (Ablank -
hydrogen atom to the synthetic free radical (28) of potassium Asample/Ablank) × 100, i.e., ratio of nonconsumed DPPH• in the
nitrosodisulfonate (radical of Fremy’s salt). A correlation presence of antioxidants to initial DPPH• in %, where Asample
between results of FRAP method and antioxidant activity is and Ablank are absorbance values of the reaction mixture with
questionable (39). Carotenoids probably are not active in the and without sample, respectively (44, 46). Other modifications
FRAP method (26, 40). (in combination with HPLC, TLC, sequential injection analysis,
The antioxidant activity of phenolic substances in the method etc.) of the method were also recommended in the literature.
increased in agreement with the number of hydroxyl groups. Unfortunately, new more or less serious advantages and
The antioxidant activity was higher for phenolic substances than disadvantages appeared in most of them, but the main principle
for ascorbic acid where only one hydroxyl group reacted at the of the basic reaction remains unchanged.
given experimental conditions. The reactivity decreased in the The amount of inactivated DPPH• is proportional to the
order gallic acid (three OH) > caffeic acid (two OH) > ferulic concentration of added flavonoids (47); thus, the classical
acid (one OH) > Trolox (one OH) > ascorbic acid (two OH) calibration procedure based on Trolox as a standard can be used
> FeSO4 in agreement with the parameters of calibration curves for quantification (48) and 1 mmol/L of Trolox corresponds to
(Table 2). Of all the used methods, the obtained absorptivity antioxidant activity of 1 mmol/L phenolic compounds.
values for standards correlated the best with the number of In our experiment, the yield of the reaction with Trolox was
phenolic hydroxyl groups (Figure 2). finished during 2 min, while with plant extracts the yield
614 J. Agric. Food Chem., Vol. 54, No. 3, 2006 Stratil et al.

Table 3. Determined Millimolar Absorptivity Coefficients and Ratio of Absolute Values of Molar Absorptivity Coefficients Related to Ferulic Acid (after
/ Symbol) of All Used Methods for Applied Standards
method gallic caffeic ferulic phenol Trolox ascorbic FeSO4
FCM 0.177/1.22 0.201/1.39 0.145 0.092/0.63 0.058/0.40 0.098/0.68 0.028/0.20
TEAC −1.588/2.54 −0.179/0.29 −0.625 −0.175/0.28 −0.253/0.40 −0.172/0.28 −0.099/0.16
FRAP 1.005/2.13 0.715/1.52 0.471 0 0.396/0.84 0.282/0.60 0.198/0.42
DPPH −1.398/5.70 −0.279/1.14 −0.244 0 −0.254/1.04 −0.147/0.60 −0.158/0.65

exponentially decreased and reached the plateau after 45-60 Table 4. Correlation of the Results of Different Methods
min (Figure 1A). The reaction rate was very fast in the
beginning (90% decrease in first 20-30 min while the rest of total free
10% in the next 30 min). The published values of antioxidant methods na rb Fc na rb Fc
activity increased with the incubation period, i.e., the values FCM vs TEAC 32 ++ ++ 31/32 ++ ++
2.2 and 3.5 were determined in the 2nd and 15th min, FCM vs FRAP 31 ++ ++ 30/32 ++ −
respectively (49). FCM vs DPPH 32 ++ ++ 30/32 +± +
The very good stability of the DPPH• radical (no measur- TEAC vs FRAP 31 ++ ++ 31/32 ++ ++
TEAC vs DPPH 31 ++ ++ 31/32 +± ++
able decrease of absorbance values was observed for DPPH• FRAP vs DPPH 31 ++ −? 31/32 +± −
reagent) was the main advantage of the method over the
TEAC method. The very long reaction times for obtaining a Number of samples. b Regression coefficient. c Paired test; ++, highly significant
plateau and thus the very long incubation times needed for correlation; +, significant correlation; −, nonsignificant correlation. Left part
measurements were the main disadvantage of the method determined by least-squares method, right half by t-test.
(similar to the other ones) as can be seen from the kinetic
data (Figure 1). The very short incubation times used in
most of the published papers were probably the main source of The determined values in the plant extracts were very similar,
the very low reproducibility of the results. Classical proce- and they correlated very well with those determined by the
dures for evaluation of the results on the basis of a suitable FRAP method (Figure 3). Because of the same basic principles
standard can be used due to the reaction kinetic and the linear of the applied methods for the determination of phenolic
relationship between absorbance and antioxidant concentration. compounds and the antioxidant activity (redox properties), one
Exactly the same initial concentrations must be used for the can expect a high correlation of determined values among all
standard and the sample solutions. Trolox was found as the methods. Correlations (Table 4) were determined by two
most suitable standard (see above) since the quantitative yields methodsslinear least-squares method (Excel) and t-test counted
of the reaction were obtained in 1 min and the values of by statistical system Unistat 4.53e. Indeed, the correlation of
decreased absorbance were similar to those of caffeic and ferulic determined values of the some extracts among all used methods
acids. (FCM-TEAC, FCM-FRAP, FCM-DPPH, TEAC-FRAP,
Hydrogen-donor capacities of polyphenols for DPPH• were TEAC-DPPH, and FRAP-DPPH) for extracts of free and total
proportional to the number of hydroxyl groups (50). The amount phenolics was highly significant (for 30-32 values, 1-2
of inactivated DPPH• was proportional to the concentration of extreme values deleted for some methods).
added flavonoids (47). The values for catechins in tea were A consensus on correlation is not general since some literature
practically equal to the number of active OH groups present in data on the contents of phenolics and their antioxidant activity
catechol and pyrogallol parts of molecules. The reactivity of are very contradictory. Some authors observed close or very
DPPH• with flavonoids not containing OH groups in the B ring close correlations (31) while the others did not find any or
and with the aromatic acids containing a single OH group is nonsignificant direct correlation. The discrepancies could be
negligible (39), but the validity of the findings probably might influenced by the estimation and interpretation of the results of
be limited. individual methods, the differences in evaluation of interferences
Ferulic acid (one OH group) was also active, and the value of other substances (like ascorbic acid, saccharides, and eventu-
was practically equal to the value for caffeic acid (two OH ally carotenoids).
groups). Five times higher values were obtained for gallic Ascorbic acid and other endiols react positively not only with
acid (three OH groups) when we compared the reactivity of the FC reagent but also with reagents of the TEAC, DPPH,
the acids tested in our experiments (Figure 2). The explanation and FRAP methods (values of ascorbic acid are equivalent to
for the lowest determined values of extracts by these methods ca. two-thirds values of Trolox applied as a standard in the
can be due to the fact that DPPH• is a long-lived little reactive methods). The corrections on the content of ascorbic acid are
radical reacting only with very reactive phenolic and other not critical and suitable for determination of overall antioxidant
antioxidants. activity and for eventual comparison of the values determined
The results were comparable for the TEAC and DPPH by different methods, since the ascorbic acid participates in the
methods despite the fact that the reactivity of some compounds total antioxidant capacity.
seriously differs. For example, the orders of decrease in However, a correction on ascorbic acid content could be
antioxidant activity determined by the TEAC (with ABTS) and applied for correlation of the total content of phenolic com-
the DPPH• methods were practically equal for seven phenolic pounds and the total antioxidant capacity. The negligible
compounds (30); however, for some other compounds, they increase of the values (<5% and mostly <1%) was observed
differed substantially (i.e., for rutin and Trolox). for the TEAC, FRAP, and DPPH methods. Interference of
Phenolic Content and Antioxidant Activity in Vegetables J. Agric. Food Chem., Vol. 54, No. 3, 2006 615

saccharides is negligible for all of the methods except for the (3) Armstrong, D., Ed. In Free Radicals and Antioxidant Protocols.
TEAC method (Table 2). Methods in Molecular Biology; Humana Press: Totowa, NJ,
In conclusion, a detailed study was performed of the contents 1998; Vol. 108, p vii.
of phenolic compounds determined by the Folin-Ciocalteu (4) Halliwell, B., Gutteridge, J. M. C., Eds. Oxygen is a toxic gas-
reagent and the antioxidant activities determined by the TEAC, an introduction to oxygen toxicity and reactive oxygen species.
In Free Radical in Biology and Medicine; Oxford University
DPPH, and FRAP methods and their correlations for extracts
Press: United Kingdom, 1999; pp 1-35.
from several species of commonly consumed vegetables. The
(5) Wallace, S. S. Oxidative damage to DNA and its repair. In
FC method for the determination of phenolic compounds is, OxidatiVe Stress and Molecular Biology of Antioxidant Defenses;
such as the methods of antioxidant activity determination, based Scandalios, J. G., Ed.; Cold Spring Harbor Laboratory Press:
on redox properties of the compounds; thus, the values could Plainview, NY, 1997; pp 49-90.
partially express the antioxidant activity. This confirms a highly (6) Spilkova, J.; Hubik, J.: Biologische wirkungen von flavonoiden.
significant correlation between the values of FC method and II. Pharm. Unserer Zeit 1992, 21, 174-182.
the values of individual methods for antioxidant activity. (7) Middleton, E., Jr.; Kandaswami, C. Efekts of flavonoids on
The three used methods for the determination of antioxidant immune and inflamatory cell function. Biochem. Pharmacol.
activity applied for the same sets of the extracts using identical 1992, 43, 1167-1179.
calibration procedures and common standard allowed the better (8) Andlauer, W.; Stehle, P.; Fürst, P. ChemopreventionsA novel
comparison of the results obtained by the TEAC, DPPH, and approach in dietetics. Curr. Opin. Clin. Nutr. Metab. Care 1998,
FRAP methods. 1, 539-547.
(9) Bors, W.; Heller, W.; Michel, Ch.; Stettmaier, K. Flavonoids
The absolute values of individual methods for determination
and polyphenols: Chemistry and biology. In Handbook of
of antioxidant activity differed proportionally. The highest values
Antioxidants; Cadenas, E., Packer, L., Eds.; Marcel Dekker: New
were obtained using the TEAC method (mean 31 of extracts York, 1996; pp 409-466.
total/free 15.3/10.1) than for the FRAP method (3.46/2.08), (10) Waterman, P. G., Mole, S., Eds. Why are phenolic compounds
while the lowest ones were obtained with the DPPH method so important? In Analysis of Phenolic Plant Metabolites. Methods
(2.90/1.75). in Ecology; Blackwell Scientific Publications: Oxford, 1994;
The simultaneous determination of ascorbic acid and the pp 44-65.
correction of absorbance values according to its contribution (11) Frankel, E. N.; Meyer, A. S. The problems of using one-
play the key role in correct determination of phenolic com- dimensional methods to evaluated multifunctional food and
pounds, mainly in vegetables rich in ascorbic acid. The biological antioxidants. J. Sci. Food Agric. 2000, 80, 1925-
determined contents of phenolic compounds could be corrected 1941.
for the ascorbic acid content to eliminate or minimize the (12) Afanas’ev, I. B.; Dorozhko, A. I.; Brodskii, A. V.; Kostyuk, A.;
misinterpretation of the ratio of real values when comparing Potapovitch, A. I. Chelating and free radical scavenging mech-
the content of phenolics and the antioxidant activity. The anism of inhibitory action of rutin and quercetin in lipid
peroxidation. Biochem. Pharmacol. 1989, 38, 1763-1769.
interference of saccharides could be more significant in samples
(13) Vinson, J. A.; Dabbagh, Y. A.; Serry, M. M.; Jang, J. Plant
rich in saccharides, but it is mostly negligible in vegetables.
flavonoids, especially tea flavonoids, are powerful antioxidants
The content of phenolic substances and total antioxidant using an in vitro oxidation model for heart disease. J. Agric.
activity correlate very well for most of the samples. There is Food Chem. 1995, 43, 2800-2802.
no significant difference in the values determined by the (14) Waterman, P. G., Mole, S., Eds. Extraction and chemical
individual methods, since the absorbance values for the methods quantification. In Analysis of Phenolic Plant Metabolites.
correspond to the absorbing substances with different molar Methods in Ecology; Blackwell Scientific Publications: Oxford,
absorptivities. More important is the common trend, i.e., the 1994; pp 66-103.
constancy of the ratio between compared values for the wider (15) Vinson, J. A.; Su, X.; Zubik, L.; Bose, P. Phenol antioxidant
set of samples. The higher deviations of the values determined quantity and quality in foods: fruits. J. Agric. Food Chem. 2001,
by the different methods for the same sample from the common 49, 5315-5321.
trend could be caused by higher content of specific phenolic (16) Vinson, J. A.; Hao, Y.; Su, X.; Zubı́k, L. Phenol antioxidant
compounds or other compounds of different reactivity (for quantity and quality in foods: Vegetables. J. Agric. Food Chem.
example, in garlic). The highest antioxidant activities were 1998, 46, 3630-3634.
(17) Wildenradt, H. L.; Singleton, V. L. The production of aldehydes
obtained for the TEAC method, substantially lower for FRAP,
as a result of oxidation of polyphenolic compounds and its
and the lowest for the DPPH method.
relation to wine aging. Am. J. Enol. Vitic. 1974, 25, 119-124.
The content of phenolic compounds and the antioxidant (18) Singleton, V. L.; Orthofer, R.; Lamuela-Raventós, R. M. Analysis
activity are partly dependent on the color of the variety of the of total phenols and other oxidation substrates and antioxidants
vegetables and total sunlight irradiation (intensity × time) during by means of Folin-Ciocalteu reagent. Methods Enzymol. 1999,
vegetative period and with the water content. The very high 299, 152-178.
values of antioxidant activity were found in intensely colored (19) Miller, N. J.; Rice-Evans, C. A.; Davies, M. J.; Gopinathan, V.;
vegetables (red cabbage, red onion, etc.), and very low values Milner, A. A novel method for measuring antioxidant capacity
were found in watery vegetables such as potato, marrow, and and its application to monitoring the antioxidant status in
cucumber. premature neonates. Clin. Sci. 1993, 84, 407-412.
(20) Lien, E. J.; Ren, S.; Bui, H.-H.; Wang, R. Quantitative structure-
activity relationship analysis of phenolic antioxidants. Free
LITERATURE CITED
Radical Biol. Med. 1999, 26, 285-294.
(1) Southorn, P. A.; Powis, G. Free radicals in medicine. 2. (21) Plumb, G. W.; Price, K. R.; Williamson, G. Antioxidant
Involvement in human-disease. Mayo Clin. Proc. 1988, 63, 390- properties of flavonol glycosides from green beans. Redox Rep.
408. 1999, 4, 123-127.
(2) Halliwell, B.; Gutteridge, J. M.; C., Cross, C. E. Free radicals, (22) Rice-Evans, C. A.; Miler, N. J.; Paganga, G. Structure-antioxidant
antioxidants, and human disease: Where are we now? J. Lab. activity relationship of flavonoids and phenolic acids. Free
Clin. Med. 1992, 119, 598-620. Radical Biol. Med. 1996, 20, 933-956.
616 J. Agric. Food Chem., Vol. 54, No. 3, 2006 Stratil et al.

(23) Re, R.; Pellegrini, N.; Proteggente, A.; Pannala, A.; Yang, M.; (38) Arts, M. J. T. J.; Dallinga, J. S.; Voss, H.-P.; Haenen, G. R. M.
Rice-Evans, C. Antioxidant activity applying an improved ABTS M.; Bast, A. A critical appraisal of the antioxidant capacity
radical cation decolorization assay. Free Radical Biol. Med. 1999, (TEAC) assay in defining optimal antioxidant structures. Food
26, 1231-1237. Chem. 2003, 80, 409-414.
(24) Long, L. H.; Halliwell, B. Antioxidant and prooxidant abilities (39) Roginsky, V.; Lissi, E. A. Review of methods to determine chain-
of food and beverages. Methods Enzymol. 2001, 335, 181-190. breaking antioxidant activity in food. Food Chem. 2005, 92,
(25) Benzie, I. F. F.; Strain, J. J. The ferric reducing ability of plasma 235-254.
(FRAP) as a measure of “antioxidant power”: The FRAP assay. (40) George, B.; Kaur, C.; Khurdiya, D. S.; Kapoor, H. C. Antioxi-
Anal. Biochem. 1996, 239, 70-76. dants in tomato (Lycopersium esculentu) as a function of
(26) Pulido, R.; Bravo, L.; Saura-Calixto, F. Antioxidant activity of genotype. Food Chem. 2004, 84, 45-51.
dietary polyphenols as determined by a modified FRAP assay. (41) Ratty, A. K.; Sunatomo, J. M.; Das, N. P. Interaction of
J. Agric. Food Chem. 2000, 48, 3396-3402. flavonoids with 1,1-diphenyl-2-picrylhydrazyl free-radical, li-
(27) Imeh, U.; Khokhar, S. Distribution of conjugated and free phenols posomal membranes and soybean lipoxygenase-1. Biochem.
in fruits: Antioxidant activity and cultivar variations. J. Agric. Pharmacol. 1988, 37, 989-995.
Food Chem. 2002, 50, 6301-6306. (42) Brand-Williams, W.; Cuvelier, M. E.; Berset, C. Use of a free
(28) Gardner, P. T.; White, T. A. C.; McPhail, D. B.; Duthie, G. G. radical method to evaluate entioxidant activity. Food Sci.
The relative contribution of vitamin C, carotenoids and phenolics Technol.-Lebensm.-Wiss. Technol. 1995, 28, 25-30.
to the antioxidant potential of fruit juices. Food Chem. 2000, (43) De Beer, D.; Jubert, E.; Wentzel, C. A.; Gelderblom, C. A.;
68, 471-474. Manley, M. Antioxidant activity of South Africa red and white
(29) Sánchez-Moreno, C.; Larrauri, J. A.; Saura-Calixto, F. A cultivar wines: Free radical scavenging. J. Agric. Food Chem.
procedure to measure the antiradical efficiency of polyphenols. 2003, 51, 902-909.
J. Sci. Food Agric. 1998, 76, 270-276. (44) Parejo, I.; Viladomat, F. M.; Bastida, J.; Rosas-Romero, A.;
(30) Kim, D. O.; Lee, K. W.; Lee, H. J.; Lee, C. Y. Vitamin C Flerlage, N.; Burillo, J.; Codina, C. Comparison between the
equivalent antioxidant capacity (VCEAC) of phenolic phy- radical scavenging activity and antioxidant activity of six distilled
tochemicals. J. Agric. Food Chem. 2002, 50, 3713-3717. and nondistilled mediterranean herbs and aroamatic plants. J.
(31) Kähkönen, M. P.; Hopia, A. I.; Vuorela, H. J.; Rauha J.-P.; Agric. Food Chem. 2002, 50, 6882-6890.
Pihlaja, K.; Kujala, T. S.; Heinonen, M. Antioxidant activity of (45) Sánchez-Moreno, C.; Larrauri, J. A.; Saura-Calixto, F. Free
plant extrakts containing phenolic compounds. J. Agric. Food radical scavenging capacity and inhibition of lipid oxidation of
Chem. 1999, 47, 3954-3962. wines, grape juices and related polyphenolic constituents. Food
(32) Karakaya, S.; El, S. N.; Tas, A. A. Antioxidant activity of some Res. Int. 1999, 32, 407-412.
foods containing phenolic compounds. Int. J. Food Sci. Nutr. (46) Tepe, B.; Sokmen, M.; Akpulat, H. A.; Sokmen, A. In vitro
2001, 52, 501-508. antioxidant activities of the methanol extract of five Allium
(33) Crozier, A.; Lean, M. E. J.; McDonald, M. S.; Black, C. species from Turkey. Food Chem. 2005, 92, 89-92.
Quantitative analysis of the flavonoid content of comercial (47) Silva, M.; Santos, M. E. R.; Caroco, C.; Rocha, R.; Justino, C.;
tomatoes, onions, lettuce, and celery. J. Agric. Food Chem. 1997, Mira, L. Structure-antioxidant activity relationships of flavonoids.
45, 590-595. Reexamination. Free Radical Res. 2002, 36, 1219-1227.
(34) Berg van den, R.; Haenen, G. R. M. M.; Berg van den, H.; Bast, (48) Russo, A.; Bonina, F.; Acquaviva, R.; Campisi, A.; Galvano,
A. Applicability of an improved Trolox equivalent antioxidant F.; Ragusa, N.; Vanella, A. Red orange extract: Effect on DNA
capacity (TEAC) assay for evaluation of antioxidant capacity cleavage. J. Food Sci. 2002, 67, 2814-2818.
measurements of mixtures. Food Chem. 1999, 66, 511-517. (49) Gaupy, P.; Dufcur, C.; Loonis, H.; Dangles, O. Quantitative
(35) Williamson, G.; Plumb, G. W.; Garcia-Conesa, M. T. Glycosy- kinetic analysis of hydrogen atom transfer reactions from dietary
lation, esterification and polymerization of flavonoids and polyphenols to the DPPH radical. J. Agric Food Chem. 2003,
hydroxycinnamates: Effects on antioxidant properties. Basic Life 51, 615-622.
Sci. 1999, 66, 483-494. (50) Bors, W.; Michel, C.; Stettmaier, K. Structure-activity relation-
(36) Berg van den, R.; Haenen, G. R. M. M.; Berg van den, H.; Vijgh ships governing antioxidant capacities of plant polyphenols.
van der, W.; Bast, A. The predictive value of the antioxidant Methods Enzymol. 2001, 335, 166-180.
capacity of structurally related flavonoids using the Trolox
equivalent antioxidant capacity (TEAC) assay. Food Chem. 2000, Received for review September 22, 2005. Revised manuscript received
70, 391-395. November 14, 2005. Accepted November 28, 2005. This work was
(37) Rezk, B. M.; Haenen, G. R. M. M.; Vijgh van der, W. J. F.; supported by a grant of the Grant Agency of the Czech Republic (GA
Bast, A. Tetrahydrofolate and 5-methyltetrahydrofolate are folates CR 521/02/1367).
with high antioxidant activity. Identification of the antioxidant
pharmacophore. FEBS Lett. 2003, 555, 601-605. JF052334J