See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/225776501

Measurements of the antioxidant capacity of fruits and vegetables using the

BR reaction method

Article  in  European Food Research and Technology · November 2002

DOI: 10.1007/s00217-002-0582-2

CITATIONS READS

35 994

2 authors:

Kerstin Höner Rinaldo Cervellati

Technische Universität Braunschweig University of Bologna
22 PUBLICATIONS   310 CITATIONS    136 PUBLICATIONS   2,864 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

video-based tool for use in science teacher education View project

Dynamics of nonlinear physicochemical and biochemical systems with modeling and predicting of their behavior under nonequilibrium conditions View project

All content following this page was uploaded by Rinaldo Cervellati on 19 December 2015.

The user has requested enhancement of the downloaded file.

Eur Food Res Technol (2002) 215:437–442
DOI 10.1007/s00217-002-0582-2

O R I G I N A L PA P E R

Kerstin Höner · Rinaldo Cervellati

Measurements of the antioxidant capacity of fruits

and vegetables using the BR reaction method

Received: 3 May 2002 / Revised: 18 June 2002 / Published online: 6 September 2002
© Springer-Verlag 2002

Abstract The antioxidant capacity of aqueous extracts lipoproteins · ORAC: Oxygen radical absorbance
of 20 vegetables and 16 fruits have been determined us- capacity · PCL: Photochemluminescence · TEAC:
ing the Briggs-Rauscher (BR) reaction method. Like oth- Trolox equivalent antioxidant capacity · TRAP: Total
er methods, the BR reaction method is also based on the radical-trapping antioxidant parameter
generation of free radicals in the reaction mixture. Anti-
oxidant scavengers of free radicals added to an active os-
cillatory BR regime cause an immediate cessation of the Introduction
oscillatory regime, an inhibition time that linearly de-
pends on the amount of the antioxidant added, and sub- There is considerable evidence that the antioxidants con-
sequent regeneration of oscillations. The BR reaction tained in fruits, vegetables and beverages play an impor-
method works at pH≈2, which is similar to that of the tant role in the maintenance of health and in prevention
fluids in the human stomach. It is known that a vegetari- of disease [1, 2, 3, 4]. A connection has been made be-
an diet can reduce the risk of stomach cancer and it is tween the intake of high carotenoid-containing fruits and
therefore interesting to determine the activity of antioxi- vegetables and protection from certain cancers [5, 6],
dants at low pH values. and recent work is also beginning to highlight the role of
Different plants were tested with the BR reaction the phenolic constituents of the diet in contributing to
method, recording potentiometrically the inhibition these protective effects. The phenolic OH group(s) con-
times produced by their extracts on an active BR mix- tained in their molecules have radical scavenging proper-
ture. The results concerning the order of the antioxidant ties and allow them to act as reducing agents, hydrogen-
activity of the examined plants are illustrated and dis- or electron-donating agents or oxygen scavengers [7, 8].
cussed. As far as the antioxidant properties of dietary agents
A comparison with the ranking order obtained with are concerned, there are situations in which knowledge
other methods is also given. of the total antioxidant potential might be more useful
than the individual polyphenolic contents [9].
Keywords Antioxidants · Briggs-Rauscher measure- Recently Schlesier et al. [10] published an assessment
ments · Fruits · Vegetables of the antioxidant activity by using six different in vitro
methods. The six common tests for measuring antioxi-
Abbreviations BR: Briggs-Rauscher · dant activity (TEAC I-III, TRAP, DPPH, DMPD, PCL
DPPH: 2,2-Diphenyl-l-picrylhydrazyl · DMPD: and FRAP) were evaluated by comparing the results of
N,N-Dimethyl-p-phenylendiamine · FRAP: Ferric four antioxidants and applying the tests to some beverag-
reducing ability of plasma · LDL: Lower density es. The assays differed in the pH (3.3–10.5) of the test-
ing system and in the nature and type of production of
K. Höner (✉) radicals. Some assays are suitable only for hydrophilic
Institut für Fachdidaktik der Naturwissenschaften, antioxidants, others for hydrophilic and lipophilic sub-
Abt. Chemie und Chemiedidaktik, stances when the solvent of the system is changed. The
Technische Universität Braunschweig, Pockelsstrasse 11,
38106 Braunschweig, Germany results showed that these six methods were not compara-
e-mail: k.hoener@tu-bs.de ble because the ranking order of the antioxidant activity
Tel.: +49 5313912876, Fax: +49 5313912845 of the examined antioxidants differed from assay to as-
R. Cervellati say. Also, different solvent systems for the same assay
Dipartimento di Chimica ‘G. Ciamician’, Università di Bologna, led to different antioxidant activities in some cases. The
Via Selmi, 2, 40126 Bologna, Italy authors concluded that this effect is probably caused by
438

different pH values. The ranking of the examined bever- total phenolic content and the BR inhibition times as
ages differed from assay to assay which led to the con- found with other methods [21].
clusion that it is only possible when comparing antioxi- The BR reaction method is based on the inhibitory ef-
dant activities to use a ranking order for each assay. The fects by antioxidant scavengers of free radicals on the
results can give an idea of the protective efficacy of sec- oscillations of the BR reaction. The generated hydrope-
ondary plant products [10]. roxyl radicals (HOO·) are among the main intermediates
A lot of studies have been made to evaluate the anti- of the BR system. The radicals produced in most of the
oxidative activities of several types of fruits and vegeta- other assays do not occur in the human body [10]. The
bles with usual methods [9, 11, 12, 13, 14, 15, 16, 17, mechanism of the action of antioxidants against HOO·
18, 19]. radicals in the BR system has recently been described
The aim of our study is to complete the information [20, 21]. When antioxidant scavengers of free radicals
about the antioxidant potential of several fruits and vege- are added to an active oscillating BR mixture there is an
tables using the BR reaction method recently reported, immediate quenching of the oscillations, an inhibition
which is based on the inhibitory effects by antioxidants time that linearly depends on the concentration of the an-
on the oscillations of the Briggs-Rauscher (BR) reaction tioxidant added in a wide range of concentration, and a
[20, 21]. The BR system consists of hydrogen peroxide, subsequent regeneration of the oscillations. Relative an-
acidic iodate, malonic acid, Mn(II) as catalyst and works tioxidant activities with respect to a substance chosen as
at pH≈2. Therefore, the BR reaction method works for a standard are determined on the basis of the inhibition
water-soluble antioxidants at a pH value that is similar to times. The inhibition time is defined as the time elapsed
that of the fluids in the human stomach. Since fruits and between the end of the addition of the antioxidant and
vegetables are like all other foods usually consumed per the first regenerated oscillation.
os, it is conceivable that they effectuate their first antiox- We have investigated the antioxidant capacities of
idant capacity against free radicals produced in the stom- aqueous extracts of several fruits and vegetables.
ach, and in this way, prevent medical problems such as
cancer to this organ [22]. The BR reaction method can
give useful in vitro information on the antioxidant activi- Experimental
ty at acidic pH values, which had been difficult prior to
the advent of this method. Materials, methods and procedure
Recently, a comparison of the results obtained from Malonic acid (Merck, reagent grade, >99%), manganese(II) sulfate
16 German white wines, using the new BR reaction monohydrate (Merck, reagent grade, >99%), NaIO3 (Merck, re-
method and the commonly used method, TEAC (trolox agent grade, >99.5%) were used without further purification.
equivalent antioxidant capacity), has been reported tak- HClO4, H2O2, and other chemicals were of analytical grade. All
ing into account the different pH values [21]. It showed stock solutions were prepared from doubly distilled, deionised
H2O. Perchloric acid was analysed by titration vs. a standard
that some antioxidants are more active under acidic con- 0.1 M NaOH solution (from Merck). H2O2 was standardised daily
ditions. This can also be interesting for the in vivo ef- by manganometric analysis.
fects of digested antioxidants. Serafini et al. [23] evalu- Oscillations in the BR mixtures were followed potentiometri-
ated the in vitro activity of green and black teas and their cally by recording the potential of a combined redox electrode
(Mettler Toledo InLab 501). The electrode was connected to a pH
in vivo effect on plasma antioxidant potential in man. multimeter (WTW, model pH 540 GLP) controlled by an IBM-
After digestion of the tea there was a prompt in vivo re- compatible PC. The accuracy of the multimeter is±1 mV. The da-
sponse, tested by the human plasma antioxidant capacity ta-acquisition program Multi Achat II (WTW) was used. The
assay. This fact suggests that the absorption of the bioac- multimeter is equipped with a temperature sensor with an accura-
tive components of tea takes place in the upper part of cy of±0.1 °C.
All solutions and reaction mixtures were maintained at con-
the gastrointestinal system, probably starting from the stant temperature (25.0 °C) using a thermostating system (accura-
stomach [23]. Also notable was the fact that the black cy±0.1 °C).
tea, which contains condensed polyphenols because of BR mixtures were prepared by mixing the appropriate amounts
the fermentation, showed a lower in vitro action than of stock solutions of reagents using pipettes or burettes in a
100 mL beaker to a total volume of 30 mL. The order of addition
green tea but an equal in vivo antioxidant action. This was: malonic acid, MnSO4, HClO4, NaIO3 and H2O2. Oscillations
led to the conclusion that the condensed polyphenols ap- started after the addition of H2O2.
pear to develop an appreciable in vivo action because of Diluted fruit or vegetable extract (1.0 mL) was added to 30 mL
structural modifications in the stomach caused by acid of an active, well-stirred BR mixture (initial composition:
[H2O2]=1.5 M, [HClO4]=0.0266 M, [IO3-]=0.0667 M, [malonic
gastric secretion, similar to the effect caused by lemon acid]=0.05 M, [Mn2+]=0.0063 M), after the third oscillation.
juice when added to a cup of black tea [23]. Therefore, it Typical potentiometric recordings of the oscillating potential of
might also be useful to utilise an assay that works at acid the solution for a non-inhibited and an inhibited BR mixture are
pH for the determination of the in vitro antioxidant activ- shown in Figs. 1a and b, respectively.
ity of food and other nutrition products. The inhibition times were then measured. The pH value of the
mixture was 1.58.
Some black and green teas have been tested with a All samples were analysed in triplicate and results were ex-
modified BR reaction method and gave similar results pressed as mean values inhibition times±standard deviation.
regarding their in vitro antioxidant activities [24]. It was
also shown that there exists a relationship between the
 439
tained (filtered extract of 200 mg vegetable in 1 mL water) was di-
luted in a suitable way with doubly distilled water.

Antioxidant potential

The antioxidant potential (AP) of fruit and vegetable extracts were

estimated on the basis of the BR inhibition times measured potent-
iometrically for different dilutions of the stock solutions as de-
scribed above. This potentiometric technique furnishes indirect
measures of the relative capacity of the antioxidants present in the
extracts to scavenge the HOO· radicals generated in the oscillating
BR mixture.

Results and discussion

The linear behaviour of the inhibition time vs. concentra-

tion of extracts added are shown in Fig. 2 and Fig. 3 for
some fruits and vegetables studied.
The parameters of the straight lines together with the
R-squared values are reported in Tables 1 a and b. It has
been discussed recently that below a certain concentration
of the extract added (different for each extract) the behav-
iour deviates from linearity. In fact, at low concentration
of extract added, the inhibition times become too low to
be measured, as shown in reference [25]. There is a
Fig. 1 a Non-inhibited reaction. Recording of the potential of a threshold under which inhibition times cannot be detected.
combined platinum electrode vs. time of an oscillating Briggs- We believe that under these lower limits, the straight lines
Rauscher (BR) mixture. b Inhibited reaction. Recording of the
potential of a combined platinum electrode vs. time when 1 mL curve towards zero. This is the reason why, in a few cases,
of a solution of caffeic acid (0.03 mg/mL) was added to 30 mL the order of activity can be reversed (see Tables 2 a and b)
of an oscillating Briggs-Rauscher (BR) mixture after the third as happens for Brussels sprout and red cabbage.
oscillation. Initial conditions: [H2O2]=1.5 mol/L, [HClO4]= The antioxidant potentials expressed by BR inhibition
0.0334 mol/L, [IO3-]=0.0667 mol/L, [malonic acid]=0.050 mol/L, times for 80 mg/mL and 100 mg/mL dilutions of extracts
[Mn2+]=0.0063 mol/L
of fruits and vegetables are presented in Tables 2 a and b.
The BR method measures the total antioxidant activi-
Preparation of the aqueous extracts ty of the aqueous extracts.
The ranking order considered is for the 100 mg/mL
Aqueous extracts of fruits extracts and can be seen directly by the order in the third
column of Table 2 a and b. In a few cases, the order dif-
The fruit extracts were prepared at room temperature because fers for the 80 mg/mL extracts because of the different
fruits normally are consumed raw. An amount of each fruit was
ground in a mortar and exactly 20 g of this material was put into a slopes of the straight lines. The carrot extract does not
measuring flask of 100 mL, which was then filled up with doubly show any appreciable activity because the main antioxi-
distilled water. The slurry (20 g/100 mL) was stirred for 10 min in dants, carotenes, are not water-soluble. The turnip ex-
the closed flask thermostated at 24 °C. Longer extraction times did tract also did not show antioxidant activity. There is a
not lead to a higher antioxidant activity. Therefore, it seems that
most of the water-soluble substances were passed in the extract. great difference between the two different types of ap-
This extract was then filtered before taking aliquots for the antiox- ples, the Braeburn apple extract being much more active
idant activity measurements. The stock solutions obtained (filtered than the Elstar. Differences are also found for different
extract of 200 mg fruit in 1 mL water) were diluted in a suitable types of other fruits or vegetables (see for example or-
way with doubly distilled water.
anges, grapes and peppers).
The assessment of the antioxidant activities of a serv-
Aqueous extracts of vegetables ing of 100 g freshly weighed fruit or vegetable shows,
for example, that the antioxidant activity of 100 g man-
The vegetable extracts were prepared under refluxed conditions darins = 135 g oranges = 254 g grapefruit = 1360 g pears
because vegetables are consumed cooked. An amount of each veg-
etable was ground in a mortar and exactly 20 g of this material and so on. All equivalences are reported in Table 3 a and
was put into a measuring flask of 100 mL, which was then filled b. The equivalences were calculated with the parameters
up with doubly distilled water. The slurry (20 g/100 mL) was of the straight line of each sample to give the same inhi-
stirred and warmed at 100 °C under reflux for 10 min. Shorter re- bition time given by 100 g of mandarins or 100 g of
fluxing times gave less antioxidant activity; longer refluxing times
did not result in higher antioxidant activity. The extract was then Brussels sprouts.
cooled down on an ice bath and filtered before taking aliquots for In the literature, there are a lot of studies in which the
the antioxidant activity measurements. The stock solution ob- antioxidant activities of fruits and vegetables have been
440
Table 1 Parameters of the straight lines and R-squared values for
the a fruits and b vegetables

Slope, m Intercept, q R2
s mL mg–1 s

a Fruit
Mandarin 44.72 –394 0.994
Orange 33.25 –313 0.998
Red orange 22.93 –67 0.999
Braeburn apple 17.77 –25 0.997
Grapefruit 16.88 –125 0.999
Pear 14.92 –243 0.997
Strawberry 7.06 +69 0.990
Elstar apple 5.31 –129 0.996
Fig. 2 Inhibition time vs. concentration of extract added for some Lychee 5.20 –141 0.994
fruits Red grape 4.12 –142 0.996
Wild strawberry 3.01 –74 0.994
Mango 2.50 –26 0.998
Banana 2.10 –67 0.979
Green grape 1.83 –93 0.999
Kiwi fruit 1.24 –40 0.988
Nectarine 1.24 –82 0.972
b Vegetables
Brussels sprout 36.55 –58 0.999
Red cabbage 32.77 +138 0.981
Garlic 12.72 –629 0.990
Black olive 11.18 –145 0.999
Bean (kidney) 8.50 +27 0.988
Button mushroom 7.84 –102 0.992
Broccoli 7.52 –156 0.997
Red pepper 6.68 –148 0.999
Spinach 5.48 –31 0.960
Green bean 5.10 –152 0.950
Corn 4.35 –22 0.994
Pea 3.14 –107 0.995
Cauliflower 2.98 –52 0.997
Onion 2.37 –81 0.986
Fig. 3 Inhibition time vs. concentration of extract added for some Celery 2.09 –71 0.995
vegetables Leek 2.08 –38 0.992
Zucchini 1.32 –38 0.992
Yellow pepper 1.07 –34 0.973
Turnip 0.52 –21 0.908
examined [8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]. Carrot 0.06 +11 0.837
Fruits and vegetables rich in anthocyanins (e.g. strawber-
ry, raspberry and red plum) demonstrated the highest an-
tioxidant activities, followed by those rich in flavanones
(e.g. orange, grapefruit) and flavones (e.g. onion, leek, major source of antioxidant activity of most fruits may
spinach, green cabbage), while the hydroxycinnamate- not be from vitamin C [14].
rich fruits (e.g. apple, tomato, pear, peach) consistently Vinson et al. [15] studied the phenol antioxidant
elicited the lower antioxidant activities [13]. Proteggente quantity and quality in vegetables and used a phenol an-
et al. [13] found that the TEAC, FRAP and ORAC val- tioxidant index, which is a combined measure of the to-
ues for each extract were relatively similar and well cor- tal phenolic content and the IC50 values obtained by the
related with the total phenolic and vitamin C contents. LDL method. They found for the wet weight measure-
The order of the TEAC values of the fresh weight ex- ments the following order: Kidney bean >garlic >onion
tracts was: strawberry = raspberry = red plum = red cab- >beet >potato >broccoli >tomato >corn >pepper >carrot
bage >>grapefruit = orange >spinach >broccoli >green >cauliflower >mushroom >celery >cabbage >lettuce
grape = onion >green cabbage >pea >apple >cauliflower, >cucumber.
pear >tomato, peach = leek >banana, lettuce. In another study, 22 fresh weight vegetables were ex-
In another study the ORAC assay was used to deter- amined using the ORAC assay but with three different
mine the antioxidant activities of several fruit with the reactive species (ROO·, HO·, Cu2+). The findings showed
following ranking: strawberry >plum >orange >red that the ranking of the antioxidant activities depend on
grape >kiwi fruit >pink grapefruit >green grape >banana the type of the reactive species in the reaction mixture
>apple >tomato >pear >melon. The order changes in [16].
some cases for the dry weight measurements. The vita- Velioglu et al. [17] found a relationship between the
min C content was also examined with the result that the total phenolic content and the antioxidant activity of
 441
Table 2 Inhibition times for 80 mg/mL and 100 mg/mL dilutions Table 3 Equivalents of a fruits
of extracts of a fruits and b vegetables in relation to 100 g of mandarin Portion (g)
and b of vegetables in relation
Inhibition time Inhibition time to 100 g of Brussels sprout a Fruit
for 80 mg/mL for 100 mg/mL Mandarin 100
s s Orange 135
Red orange 184
a Fruit Braeburn apple 247
Mandarin 3079±46 4163±63 Grapefruit 254
Orange 2288±34 3067±46 Pear 295
Red orange 1777±27 2236±34 Strawberry 580
Braeburn apple 1228±19 1570±24 Elstar apple 811
Grapefruit 1226±19 1561±23 Lychee 825
Pear 978±15 1265±19 Wild strawberry 1408
Strawberry 609±10 797±12 Mango 1676
Elstar apple 252±4 391±6 Banana 2014
Lychee 274±4 388±6 Green grape 2326
Wild strawberry 195±3 245±4 Kiwi fruit 3390
Red grape 180±3 273±4 Nectarine 3423
Mango 169±3 222±4 b Vegetable
Banana 105±2 126±3
Kiwi fruit 58±2 84±3 Brussels sprout 100
Green grape 54±2 92±3 Red cabbage 105
Nectarine 18±1 49±2 Black olive 335
Garlic 335
b Vegetables Bean (kidney) 420
Red cabbage 2899±44 3563±53 Button mushroom 472
Brussels sprout 2861±43 3600±54 Broccoli 500
Black olive 742±11 968±15 Red pepper 561
Bean (kidney) 661±10 890±14 Spinach 663
Button mushroom 545±9 703±11 Green bean 736
Spinach 469±7 511±8 Corn 833
Broccoli 434±7 581±9 Pea 1181
Garlic 420±6 588±9 Cauliflower 1226
Red pepper 377±6 526±8 Onion 1554
Corn 310±5 419±7 Leek 1749
Green bean 254±4 302±5 Celery 1757
Cauliflower 178±3 244±4 Zucchini 2756
Pea 138±2 199±3 Yellow pepper 3396
Onion 94±2 150±3 Turnip 6963
Celery 88±1 138±2
Leek 74±1 116±2
Zucchini 61±1 95±2
Yellow pepper 48±1 74±2
Carrot 15±1 17±1 For the study of antioxidative activity, some factors of
Turnip 15±1 16±1 the method have to be considered i.e. the practicability,
instrumental requirements and the time, expertise and
cost necessary for the analysis. Concerning these as-
pects, the BR method reaction described and used here
plant materials using the β-carotene bleaching method. has many advantages. The analysis is inexpensive and
This was also found in other studies [21, 26, 27]. rapid and the reagents and apparatus are commonly used
However, despite these differences, the BR reaction in all chemical laboratories.
method also gives an idea of the protective efficacy like
all the other in vitro assays. The differences found in the
ranking order of the antioxidative activity in comparison References
to other assays surely depends on the different pH and 1. Ames B, Shigenaga MK, Hagen TM (1993) Proceedings of
the different radicals formed in the reaction mixture. the National Academy of Sciences of the United States of
From the experimental results and the data treatment America 90:7915–7922
presented here, the oscillating BR reaction is suitable as 2. Kinsella JE, Frankel E, German B, Kanner J (1993) Food Tech
an analytical method to measure the relative in vitro anti- 47:85–89
3. Halliwell B, Antonia Murcia M, Chirico S, Aruoma I (1995)
oxidative potential of fruits and vegetables at a low pH Critical Rev Food Science Nutr 35:7–20
value, similar to the pH in the stomach. It has recently 4. Halliwell B (1997) Nutr Rev 55(1):S44-S52
been shown for some pure antioxidants that the inhibi- 5. Block G (1992) Nutr Rev 50:207–213
tion time linearly depends on the antioxidant concentra- 6. Block G, Patterson B, Subar A (1992) Nutr Cancer 18:1–29
7. Kandaswami C, Middleton E (1994) Adv Exp Med Biol 366:
tion between 0.001 and 0.16 mM in mixture [20]. There- 351–376
fore, the sensitivity of the BR assay is comparable with 8. Kinsella JE, Frankel E, German B, Kanner J (1993) Food
those of other methods. Technol 47:85–89
442
9. Paganda G, Miller N, Rice-Evans CA (1999) Free Radic Res 19. Kahkonen MP, Hopia AI, Vuorela HJ, Rauha J-P, Pihlaja K,
30:153–162 Kujala TS, Heinonen M (1999) J Agric Food Chem 47:
10. Schlesier K, Harvat M, Böhm V, Bitsch R (2002) Free Radic 3954–3962
Res 36:177–187 20. Cervellati R, Höner K, Neddens C, Costa S (2001) Helv Chim
11. Cao G, Sofic E, Prior RL (1996) J Agric Food Chem 44: Acta 84:3533–3547
3426–3431 21. Höner K, Cervellati R, Neddens C (2002) Eur Food Res Tech-
12. Wang H, Cao G, Prior RL (1996) JAgric Food Chem 44: nol 214:356–360
701–705 22. Dekant W, Vamvakas S (1995) Toxikologie. Spektrum Akade-
13. Proteggente AR, Pannala AS, Paganga G, van Buren L, Wag- mischer Verlag, Heidelberg
ner E, Wisemann S, van de Put F, Dacombe C, Rice-Evans CA 23. Serafini M, Ghiselli A, Ferro-Luzzi A (1996) Eur J Clin Nutr
(2002) Free Rad Res 36:217–233 50:28–32
14. Wang H, Cao G, Prior RL (1996) J Agric Food Chem 44: 24. Höner K, Cervellati R (2002) Pharm Educ 1:37–42
701–705 25. Cervellati R, Crespi-Perellino N, Furrow SD, Minghetti A
15. Vinson JA, Hao Y, Su X, Zubik L (1998) J Agric Food Chem (2000) Helv Chim Acta 83:3179–3190
46:3630–3634 26. Baderschneider B (2000) Isolierung und Strukturaufklärung
16. Cao G, Sofic E, Prior RL (1996) J Agric Food Chem 44: antioxidativ wirksamer Verbindungen aus Weißwein. Disserta-
3426–3431 tion, Technische Universität Braunschweig
17. Velioglu YS, Mazza G, Gao L, Oomah BD (1998) J Agric 27. Baderschneider B, Luthria D, Waterhouse AL, Winterhalter P
Food Chem 46:4113–4117 (1999) Vitis 38:127–131
18. Hertog MGL, Fleuriet A, Billot J (1992) J Agric Food Chem
40:2379–2383

View publication stats