Analogs of anthocyanins with a 3′,4′-dihydroxy substitution: Synthesis and investigation of their acid-base, hydration, metal binding and hydrogen-donating properties in aqueous solution

Article abstract of DOI:10.1016/j.dyepig.2012.07.006

Glycosides of hydroxylated flavylium ions are proposed as pertinent analogs of anthocyanins, a major class of polyphenolic plant pigments. Anthocyanins with a 3′,4′-dihydroxy substitution on the B-ring (catechol nucleus) are especially important for their metal chelating and electron-donating (antioxidant) capacities. In this work, an efficient chemical synthesis of 3′,4′-dihydroxy-7-O-β-d-glucopyranosyloxyflavylium chloride and its aglycone is reported. Then, the ability of the two pigments to undergo proton transfer (formation of colored quinonoid bases) and add water (formation of a colorless chalcone) is investigated: at equilibrium the colored quinonoid bases (kinetic products) are present in very minor concentrations (<10% of the total pigment concentration) compared to the colorless chalcone (thermodynamic product). The glucopyranosyloxyflavylium ion appears significantly less acidic than the aglycone. The thermodynamics of the overall sequence of flavylium - chalcone conversion is not affected by the β-d-glucosyl moiety while the kinetics appears slower by a factor ca. 8. Although the glucopyranosyloxyflavylium ion and its aglycone display similar affinities for Al³⁺, the Al³⁺-glucoside complex is more stable than the Al³⁺-aglycone complex due to the higher sensitivity of the latter to water addition and conversion into the corresponding chalcone. Finally, the glucopyranosyloxyflavylium ion and its aglycone are compared for their ability to reduce the 1,1-diphenyl-2-picrylhydrazyl radical in a mildly acidic water/MeOH (1:1) mixture as a first evaluation of their antioxidant activity. Glycosidation at C₇-OH results in a lower rate constant of first electron transfer to DPPH and a lower stoichiometry (total number of 1,1-diphenyl-2-picrylhydrazyl radicals reduced per pigment molecule). Anthocyanins are difficult to extract from plants in substantial amount. However, the analogs investigated in this work are of easy access by chemical synthesis and express the physico-chemical properties typical of anthocyanins. They can thus be regarded as valuable models for investigating the coloring, metal-binding and antioxidant properties of these important natural pigments.

Full text of DOI:10.1016/j.dyepig.2012.07.006

Dyes and Pigments 96 (2013) 7e15
Contents lists available at SciVerse ScienceDirect
Dyes and Pigments
journal homepage: www.elsevier.com/locate/dyepig
Analogs of anthocyanins with a 30,40-dihydroxy substitution: Synthesis and
investigation of their acidebase, hydration, metal binding and hydrogen-donating
properties in aqueous solution
*
Nathalie Mora-Soumille, Sheiraz Al Bittar, Maxence Rosa, Olivier Dangles
University of Avignon, INRA, UMR408, 84000 Avignon, France
a r t i c l e i n f o
a b s t r a c t
Article history:
Glycosides of hydroxylated flavylium ions are proposed as pertinent analogs of anthocyanins, a major
class of polyphenolic plant pigments. Anthocyanins with a 30,40-dihydroxy substitution on the B-ring
(catechol nucleus) are especially important for their metal chelating and electron-donating (antioxidant)
Received 19 May 2012
Received in revised form
3 July 2012
Accepted 9 July 2012
Available online 20 July 2012
capacities. In this work, an efficient chemical synthesis of 30,40-dihydroxy-7-O-
b-D-glucopyranosyloxy-
flavylium chloride and its aglycone is reported. Then, the ability of the two pigments to undergo proton
transfer (formation of colored quinonoid bases) and add water (formation of a colorless chalcone) is
investigated: at equilibrium the colored quinonoid bases (kinetic products) are present in very minor
concentrations (<10% of the total pigment concentration) compared to the colorless chalcone (ther-
modynamic product). The glucopyranosyloxyflavylium ion appears significantly less acidic than the
aglycone. The thermodynamics of the overall sequence of flavylium e chalcone conversion is not affected
Keywords:
Anthocyanin
Flavylium
Glycoside
by the b-D-glucosyl moiety while the kinetics appears slower by a factor ca. 8. Although the glucopyr-
Synthesis
anosyloxyflavylium ion and its aglycone display similar affinities for Al3þ, the Al3þ-glucoside complex is
more stable than the Al3þ-aglycone complex due to the higher sensitivity of the latter to water addition
and conversion into the corresponding chalcone. Finally, the glucopyranosyloxyflavylium ion and its
aglycone are compared for their ability to reduce the 1,1-diphenyl-2-picrylhydrazyl radical in a mildly
acidic water/MeOH (1:1) mixture as a first evaluation of their antioxidant activity. Glycosidation at C7-OH
results in a lower rate constant of first electron transfer to DPPH and a lower stoichiometry (total number
of 1,1-diphenyl-2-picrylhydrazyl radicals reduced per pigment molecule).
Aluminum
Antioxidant
Anthocyanins are difficult to extract from plants in substantial amount. However, the analogs inves-
tigated in this work are of easy access by chemical synthesis and express the physico-chemical properties
typical of anthocyanins. They can thus be regarded as valuable models for investigating the coloring,
metal-binding and antioxidant properties of these important natural pigments.
Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction
humans [3] severely restrict the biological significance of in vitro
antioxidant tests, it is reasonable to assume that polyphenols with
Polyphenols with a 1,2-dihydroxybenzene (catechol) group are
common in plants and in our diet. This is for instance the case of
caffeic acid (3,4-dihydroxycinnamic acid) and 30,40-dihydroxy-
flavonoids such as quercetin, (epi)catechin, cyanidin and their
derivatives (O-glycosides, esters, oligomers) [1,2]. Those poly-
phenols are of special interest for their ability to bind metal ions
and readily transfer electrons or H-atoms to radicals. As such, they
are typically strong in vitro antioxidants. Although the relatively
poor bioavailability and extensive metabolism of polyphenols in
a catechol group may be very important antioxidants in plant and
food, and possibly in the digestive tract [4,5].
Anthocyanins are naturally occurring glycosides of flavylium (2-
phenyl-1-benzopyrylium) ions substituted by hydroxyl and
methoxyl groups [6]. As most polyphenols, they are mainly stored
in the mildly acidic aqueous environment of vacuoles within plant
cells. Anthocyanins constitute one of the major classes of plant
pigments, typically responsible for a wide variety of red to blue
colors [7]. One important mechanism of color stabilization and
variation is metal e anthocyanin binding, a phenomenon restricted
to 30,40-dihydroxyflavylium ions. In particular, most blue colors
* Corresponding author. Tel.: þ33 490 14 44 46; fax: þ33 490 14 44 41.
E-mail address: olivier.dangles@univ-avignon.fr (O. Dangles).
0143-7208/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.dyepig.2012.07.006

8
N. Mora-Soumille et al. / Dyes and Pigments 96 (2013) 7e15
found in Nature are complexes of anthocyanins with magnesium,
aluminum or iron ions. On the other hand, anthocyanins are also
one of the most ubiquitous polyphenol classes in foods, e.g. berries
and their products (juices, jams, red wine), red cabbage, red onion
and eggplant [2]. As such, anthocyanins may contribute to the
protective effects of diets rich in plant products [8]. However,
investigating the chemical properties of anthocyanins in line with
their activity in plant, food and in humans is somewhat impeded by
their difficult extraction and purification from plants and their
limited access from commercial sources. Hence glycosides of
hydroxylated flavylium ions can be proposed as pertinent antho-
cyanin analogs. In this work, an efficient chemical synthesis of 30,40-
2.2.2. 30,40,7-Trihydroxyflavylium chloride (P1)
solution of equimolar amounts (4 mmol) of 2,4-
A
dihydroxybenzaldehyde and 3,4-dihydroxyacetophenone in
distilled EtOAc (10 mL) was cooled to 0 ꢁC. Gaseous HCl (generated
by action of 98% H2SO4 on solid NaCl) was gently bubbled through
the solution for 90 min. The mixture was kept at 4 ꢁC for 3 days,
then filtered. More precipitate was collected after evaporation of
the filtrate and addition of diethyl ether (Et2O). After precipitation
in EtOAc, P1 was obtained as a red powder (0.651 g, yield 56%). The
purity of P1 was carefully checked by reversed-phase HPLC. 1H
0
NMR [0.2 M TFA-d in CD3OD]:
d
¼ 7.08 (1H, d, J ¼ 8.8, H5 ), 7.40 (1H,
dd, J ¼ 2.2 and 8.8, H6), 7.46 (1H, d, J ¼ 2.2, H8), 7.84 (1H, d, J ¼ 2.2,
0
0
dihydroxy-7-O-
b
-D-glucopyranosyloxyflavylium chloride and its
H2 ), 8.00 (1H, dd, J ¼ 2.2 and 8.8, H6 ), 8.11 (1H, d, J ¼ 8.8, H5), 8.24
aglycone is reported as well as the ability of the two pigments to
undergo proton transfer, add water with subsequent conversion
into chalcones, bind Al3þ and deliver electrons to the DPPH (1,1-
diphenyl-2-picrylhydrazyl) radical.
(1H, d, J ¼ 8.8, H3), 8.99 (1H, d, J ¼ 8.8, H4). 13C NMR [0.2 M TFA-d in
0
0
CD3OD]:
d
0
¼ 103.0 (C8), 112.9 (C3), 115.7 (C2 ), 117.3 (C5 ), 119.0 (C10),
0
0
121.0 (C1 ), 121.5 (C6), 125.7 (C6 ), 133.1 (C5), 147.8 (C3 ), 153.1 (C4),
0
156.7 (C4 ), 159.4 (C9), 169.3 (C7), 173.1 (C2). HRMS m/z ¼ 255.0652
(Mþ, 255.0652 calculated for C15H11Oþ4 ). UV/Vis (0.13 M aqueous
HCl):
lmax ¼ 472 nm.
3
(470 nm) ¼ 38,400 Mꢂ1 cmꢂ1. HPLC-UV/Vis tR ¼ 24.1 min,
2. Experimental
2.1. Materials and instruments
2.2.3. 4-(20,30,40,60-Tetra-O-acetyl-
hydroxybenzaldehyde
b-
D-glucopyranosyloxy)-2-
All starting materials were obtained from commercial suppliers
A solution of tetra-O-acetyl-a-D-glucopyranosylbromide (9.25 g,
and were used without purification. Purifications were performed
by column chromatography on Merck Si60 silica gel (40e63 mm)
1.5 equiv.) in CH2Cl2 (25 mL) was added to a solution of 2,4-
dihydroxybenzaldehyde (2.07 g, 15 mmol) and tris(2-(2-
methoxyethoxy)ethyl)amine (7.20 mL, 1.5 equiv.) in 1 M NaHCO3/
1 M KCl (25 mL, 1/1, v/v). The mixture was refluxed for 48 h. After
addition of H2O (100 mL) and extraction with CH2Cl2 (3 ꢀ 100 mL),
the combined organic phases were successively washed with 1 M
HCl (2 ꢀ 100 mL) and H2O (2 ꢀ 100 mL), dried over Na2SO4 and
concentrated. The residue was purified on silica gel (eluent EtOAc/
cyclohexane (3/7, v/v)) to afford the target compound as a white
and by elution on Varian bond elut C18 silica gel cartridges.
1H and 13C NMR spectra were recorded on an Advance DPX300
Bruker apparatus at 300.13 MHz (1H) or 75.46 MHz (13C). Chemical
shifts (d
) in ppm relative to tetramethylsilane, 1He1H coupling
constants (J) in Hz. High-resolution mass spectrometry (HRMS)
analyses were carried out on Qstar Elite instrument (Applied Bio-
systems SCIEX). Mass detection was performed in the positive
electrospray ionization mode. HPLC analyses were performed on
a Waters HPLC system consisting of a 600E pump, a 717 autosam-
pler, a 2996 photodiode array detector, an in-line AF degasser and
a Millenium workstation. A LichroCart 250-4 Lichrospher 100
powder (5.62 g, yield 80%). 1H NMR [CDCl3]:
d
¼ 2.08 (12H, s, 4 CH3
of Ac groups), 3.95 (1H, m, H5’), 4.17e4.33 (2H, 2dd, J ¼ 12.4 and 5.8,
0
0
0
0
0
J ¼ 12.4 and 2.3, H6 ), 5.14e5.32 (4H, m, H1 , H2 , H3 , H4 ), 6.54 (1H, s,
H3), 6.59 (1H, d, J ¼ 8.5, H5), 7.47 (1H, d, J ¼ 8.5, H6). 13C NMR
0
0
0
RP18e column (250 ꢀ 4.6 mm, 5
mm particle size) was used for
[CDCl3]:
d
¼ 20.6 (4 CH3 of Ac groups), 61.8 (C6 ), 68.1 (C4 ), 70.8 (C2 ),
chromatographic separations at 25 ꢁC. The solvent system was
a gradient of A (5% HCO2H in MeCN/H2O 1/1) and B (5% HCO2H in
72.3 (C5 ), 72.5 (C3 ), 97.6 (C1 ), 103.5 (C3), 109.6 (C5), 116.6 (C6), 135.4
(C1), 163.1 (C2), 164.0 (C4), 169.2e170.6 (4 C]O of Ac groups), 194.9
(CHO).
0
0
0
H2O) with 10%
A at 0 min and 100% A at 60 min (flow
rate ¼ 1 mL minꢂ1). UVeVis absorption spectra were recorded on
an Agilent 8453 diode array spectrometer equipped with
a magnetically stirred quartz cell (optical path length ¼ 1 cm). The
temperature in the cell was controlled by means of a water-
thermostated bath at 25 ꢃ 0.1 ꢁC.
2.2.4. 30,40-Dihydroxy-7-O-
b-D-glucopyranosyloxyflavylium
chloride (P2)
Equimolar amounts (1 mmol) of 3,4-dihydroxyacetophenone
and -glucopyranosyloxy)-2-
4-(20,30,40,60-tetra-O-acetyl-
b-D
hydroxybenzaldehyde were dissolved in dry EtOAc (10 mL) and
cooled to 0 ꢁC. Gaseous HCl was gently bubbled through the solu-
tion under stirring during 60 min. The deep-red solution was then
allowed to stay at ꢂ18 ꢁC for 6 days and filtered. Et2O was added to
the filtrate to ensure complete precipitation. The solid was dis-
solved in MeOH (20 mL) under Ar and a solution of MeONa in
MeOH was added until pH 9 (wet pH paper). After stirring for 1.5 h
at room temperature, 1 M HCl was added until pH 1 (wet pH paper).
The mixture was kept at 4 ꢁC for 12 h, then concentrated under
reduced pressure. The residue was dissolved in 0.01 M HCl (2 mL)
and loaded on a C18 cartridge. After elution with 100 mL of 0.01 M
HCl to remove contaminating NaCl, P2 was eluted with 70 mL of
0.2 M HCl in MeOH. After evaporation of solvent under reduced
pressure, P2 was obtained as a red powder (0.34 g, yield 75%). The
purity of P2 was carefully checked by reversed-phase HPLC. 1H
00
2.2. Chemical syntheses
2.2.1. 3,4-Dihydroxyacetophenone
A mixture of activated zinc powder (5 g, 76 mmol),
u-chloro-
3,4-dihydroxyacetophenone (5 g, 27 mmol), THF (120 mL) and
acetic acid (30 mL) was vigorously stirred for 2 days at room
temperature. After filtration and concentration under reduced
pressure, EtOAc (100 mL) was added. The organic layer was washed
with water (3 ꢀ 100 mL), dried over Na2SO4 and evaporated. The
crude product was purified by column chromatography (SiO2,
cHex/EtOAc,
dihydroxyacetophenone as a white amorphous powder (3.7 g).
Yield 90%. 1H NMR [CDCl3]:
¼ 2.53 (s, 3H, COCH3), 5.99 (1H, s, OH),
1:1
v/v)
to
give
compound
3,4-
d
6.19 (1H, s, OH), 6.96 (1H, d, J ¼ 8.3, H5), 7.55 (1H, dd, J ¼ 2.0 and 8.3,
NMR [0.2 M TFA-d in CD3OD]:
00 00 00 00
3.55e3.61 (2H, m, H3 , H2 ), 3.69e3.79 (2H, m, H5 , H6 ), 3.99 (1H,
d
¼ 3.47 (1H, t, J ¼ 9.2, H4 ),
H6), 7.67 (1H, d, J ¼ 2.0, H2). 13C NMR [CDCl3]:
d
¼ 24.9 (CH3), 114.4,
00
00
0
114.6 (C1, C5), 122.2 (C2), 129.2 (C6), 145.0 (C3), 150.9 (C4), 198.4 (C]
O). HPLC-UV/Vis tR ¼ 14.2 min, lmax ¼ 276 nm.
m, H6 ), 5.37 (1H, d, J ¼ 6.7, H1 ), 7.11 (1H, d, J ¼ 8.8, H5 ), 7.63 (1H,
dd, J ¼ 2.2 and 8.9, H6), 7.91 (1H, d, J ¼ 2.2, H8), 7.95 (1H, d, J ¼ 2.2,

N. Mora-Soumille et al. / Dyes and Pigments 96 (2013) 7e15
9
0
H2’), 8.12 (1H, dd, J ¼ 2.2 and 8.8, H6 ), 8.24 (1H, d, J ¼ 8.9, H5), 8.42
2.4. Data analysis
(1H, d, J ¼ 8.9, H3), 9.08 (1H, d, J ¼ 8.9, H4). 13C NMR [0.2 M TFA-d in
00
00
00
00
00
CD3OD]:
d
¼ 62.5 (C6 ), 71.2 (C4 ), 74.6 (C2 ), 77.8 (C5 ), 78.7 (C3 ),
All calculations were carried with the Scientist program
(MicroMath, Salt Lake City, USA). Curve-fittings were achieved
through least square regression and yielded optimized values for
the parameters (rate constants, thermodynamic constants, molar
absorption coefficients, stoichiometries, see eqs in text). Standard
deviations are reported. Good (>0.99) to excellent (>0.999)
correlation coefficients were typically obtained.
00
0
0
101.9 (C1 ), 105.0 (C8), 109.3 (C3), 115.5 (C2 ), 116.9 (C5 ), 118.3 (C10),
0
0
0
121.6 (C1 ), 122.3 (C6), 127.5 (C6 ), 133.0 (C5), 148.7 (C3 ), 153.8 (C4),
0
158.6 (C4 ), 159.2 (C9), 167.0 (C7), 175.0 (C2). HRMS m/z ¼ 417.1178
(Mþ, 417.1180 calculated for C21H21Oþ9 ). UV/Vis (0.13 M aqueous
HCl):
3
(466 nm) ¼ 14300 Mꢂ1 cmꢂ1. HPLC-UV/Vis tR ¼ 15.3 min,
lmax ¼ 467 nm.
3. Results and discussion
2.3. Spectroscopic measurements
3.1. Synthesis of pigments
2.3.1. Proton transfer and hydration reactions
To 2 mL of a 0.05 M acetate buffer (pH 3e6) containing 0.5 M
NaCl and placed in the spectrometer cell was added a small volume
The route for the synthesis of 30,40-dihydroxy-7-O-
pyranosyloxyflavylium chloride (P2) was adapted from one of our
previous work [9] reporting the synthesis of 30,7-dihydroxy-40-O-
-glucopyranosyloxyflavylium chloride (Scheme 1). Due to the
strong hydrogen bond between the carbonyl group and C2-OH of
2,4-dihydroxybenzaldehyde, the latter group is probably much less
acidic than C4-OH. Consequently, glycosidation under mildly alka-
line phase transfer conditions regioselectively took place at C4-OH.
The flavylium chromophore was then constructed via acid-
catalyzed aldol condensation followed by cyclization and subse-
quent dehydration. Methanolysis of the acetate protecting groups,
acidification and purification afforded pigment 2 in good yield.
b-D-gluco-
(20e40
m
L) of a concentrated solution of pigment in acidified MeOH
M. A
b
-
(0.2 M HCl). The pigment concentration in the cell was 50
m
D
first UVevisible spectrum was recorded immediately for pKa
determination. A second UVevisible spectrum was recorded after
overnight equilibration in the dark for pK0h determination. For
investigating the kinetics of flavylium e chalcone conversion,
UVevisible spectra were recorded at regular time intervals (90 s)
over 2 h following pigment addition. Each experiment was
repeated three times at different pH.
2.3.2. Aluminum complexation
To 2 mL of a 0.1 M acetate buffer (pH 4) placed in the spec-
trometer cell were successively added 50
(prepared from Al2(SO4)3, 18H2O, concentration range: 1e10 mM)
in 0.05 M aqueous HCl and 50 L of a freshly prepared 2 mM
pigment solution in acidified MeOH (0.1 M HCl). Spectra were
typically recorded every 0.5 s over 1 min or every 15 s over 15 min.
Spectra were also recorded after overnight equilibration.
m
L of Al3þ solution
3.2. Structural transformations in water
m
When the pH is increased from 2 to 6, flavylium ions (AHþ) are
typically converted into neutral quinonoid bases (A, kinetic prod-
ucts) by proton transfer (thermodynamic constant Ka) and in
a mixture of colorless forms (thermodynamic products) by the
following sequence (Scheme 2): water addition at C2 yielding
hemiketal B (thermodynamic constant Kh), cycle-chain tautomeri-
zation of B into (Z)-chalcone CZ (thermodynamic constant Kt) and
isomerization of CZ into (E)-chalcone CE (thermodynamic constant
Ki) [6,10,11]. Due to their planarity and extensive electron delocal-
ization over the 3 rings, flavylium ions lacking an O-glycosyloxy
group at C3 are less prone to water addition than natural antho-
cyanins. Consequently, the flavylium ions of P1 and P2 are slowly
converted into CE with no significant accumulation of hemiketal B
2.3.3. Hydrogen abstraction by DPPH
To 1 mL of a freshly prepared 0.2 mM solution of DPPH in MeOH
was added 1 mL of 0.1 M acetate buffer (pH 3.5) in the spectrometer
cell (final pH ca. 4.4). Aliquots (20e60 mL) of a freshly prepared
0.1 mM solution of the antioxidant in acidified MeOH (0.01 M HCl).
Spectra were recorded every 0.5 s over 2 min for the determination
of rate constants and partial stoichiometries. Kinetic runs over
60 min were used for the determination of total stoichiometries.
OH
OH
OH
Cl
OH
HO
O
i
+
OH
HO
P1
O
O
H
AcO
AcO
OAc
O
ii
AcO
HO
OH
O
OH
OH
4'
3'
Cl
OH
OH
iii
6"
B
O
+
O
O
1"
OH
2
7
HO
HO
A
C
P2
O
O
H
4
5
Scheme 1. Chemical synthesis of P1 and P2: (i) Gaseous HCl, AcOEt. (ii) tetra-O-acetyl-a-D-glucopyranosylbromide, tris(2-(2-methoxyethoxy)ethyl)amine in CH2Cl2/H2O (1 M
NaHCO3, 1 M KCl). (iii) Gaseous HCl, AcOEt, then MeONa, MeOH, then aq. HCl.

10
N. Mora-Soumille et al. / Dyes and Pigments 96 (2013) 7e15
Scheme 2. Structural transformations of 30,40,7-trihydroxyflavylium ion (P1) in mildly acidic aqueous solution.
and CZ. This particular behavior makes it possible to independently
investigate proton transfer and chalcone formation. Thus, addition
of a strongly acidic concentrated solution of P1 or P2 in MeOH to
aqueous buffers at pH 3e6 with immediate recording of the
UVevisible spectra permits the determination of the thermody-
namic constant of proton transfer (Ka ¼ [Hþ][A]/[AHþ]) while the
UVevisible spectra recorded after overnight equilibration give
access to the overall thermodynamic constant of water addition
(K0h ¼ KhKtKi ¼ [Hþ][CE]/[AHþ]). The typical UVevisible spectra of
AHþ, A and CE are shown in Fig. 1.
Eqs. (1) and (2) were used in the curve-fitting analyses for pKa
and pK0h determination, respectively (Fig. 2, Table 1). Combining
eqs. (1) and (2) gives eq. (3), which can also be used for the esti-
mation of K0h from the amplitude
D
A ¼ A0 ꢂ Af of the time-
dependence of the visible absorbance during water addition
(Fig. 3, Table 2).
ꢀ
ꢁ
ꢀ
ꢁ
A0 ꢂ Af
D
A
Kh0
¼
10ꢂpH þ Ka
¼
10ꢂpH þ Ka
(3)
Af
A0 ꢂ
D
A
Combining the Beer’s law, the expressions of the thermody-
namic constants Ka and K0h and the equation of pigment conser-
vation readily permits the derivation of the changes in visible
absorbance as a function of pH for freshly prepared pigment solu-
tion (eq. (1), water addition neglected) and solutions equilibrated
overnight (eq. (2)).
Unexpectedly, with both P1 and P2, when the observed rate
constant of flavylium hydration kohbs is plotted as a function of the
proton concentration, a monotonous increase is observed (Fig. 4),
which is in sharp contrast with the usual behavior [6]. Indeed,
quinonoid bases are less electrophilic than the flavylium ion and do
not undergo water addition in mildly acidic solution. Hence, khobs
normally decays when the pH increases as a consequence of less
flavylium ion and more quinonoid bases being present in solution.
One has: kohbs ¼ k0hxAH þ k0ꢂh[Hþ] (k0h: overall rate constant for the
flavylium to chalcone conversion, k0ꢂh: overall rate constant for the
chalcone to flavylium conversion) with xAH þ xA ¼ 1, Ka ¼ [Hþ]xA/
xAH. Combining these equations yields eq. (4).
1 þ rAKa10pH
A0 ¼ AAH
Af ¼ AAH
(1)
(2)
1 þ Ka10pH
1 þ rAKa10pH
À
Á
1 þ Ka þ Kh0 10pH
AAH
¼
3 AHC (absorbance of a strongly acidic pigment solution
kh0
Â
Ã
kohbs
¼
þ k0ꢂh Hþ
(4)
containing the sole flavylium ion, C: total pigment concentration),
rA AH (ratio of the molar absorption coefficients of quinonoid
bases and flavylium ion).
Ka
¼
3
A
/
3
Â
Ã
1 þ
Hþ

N. Mora-Soumille et al. / Dyes and Pigments 96 (2013) 7e15
11
A
0.7
A(520 nm)
1.2
1
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
0.8
0.6
0.4
0.2
0.0
5
300
400
500
600
700
800
2
3
4
5
6
pH
A
0.7
A(470 nm)
2.0
1
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.6
1.2
0.8
0.4
4
300
400
500
600
700
800
0.0
2
3
4
5
6
Fig. 1. pH-dependence of the UVevisible spectra of P2. Top: Spectral measurements
immediately after addition of pigment to buffer (1: pH 2.0, 2: pH 4.0, 3: pH 4.4, 4: pH
5.0, 5: pH 6.0). Down: Spectral measurements on solutions equilibrated overnight (1:
pH 2.0, 2: pH 3.0, 3: pH 3.5, 4: pH 4.0).
pH
Fig. 2. pH dependence of the visible absorbance of P1. Top: Spectral measurements
immediately after addition of pigment to buffer (the solid line is the result of the
curve-fitting according to eq. (1)). Down: Spectral measurements on solutions equili-
brated overnight the solid line is the result of the curve-fitting according to eq. (2).
Obviously, eq. (4) does not hold with P1 and P2. Unlike natural
anthocyanins, those flavylium ions are strongly conjugated as no
substituent at C3 restricts the planarity of the 3 rings. Hence,
pK0h from fully equilibrated solutions. This instability may reflect its
sensitivity to autoxidation (initiated by unidentified metal traces).
The close pK0h values deduced from the amplitude of the time-
dependence of the visible absorbance during water addition
(Table 2) show that the thermodynamics of the overall hydration-ring
0
conjugation of O4 and O7 with the pyrylium C-ring must be very
strong and this may lower the positive charge at C2 to the point that
hydration only takes place via the minor hydroxide ion. If so, the
observed rate constant of flavylium hydration can be rewritten as:
kohbs ¼ k0hxAH[HOꢂ] þ k0ꢂh. Consequently, eq. (4) is changed into eq.
(5) (Kw: ionic product of water):
opening-(Z,E) isomerization process is not affected by the
b-D-
Table 1
k0hKw
Thermodynamic constants of water addition and proton transfer.
kohbs
¼
þ k0ꢂh
(5)
Â
Ã
Hþ þ Ka
Pigment
pKobs
AAH, rA
¼
3
/
3
l (nm), r
(number of points)
A
AH
Eq. (5) actually permits a good curve-fitting of kohbs vs. proton
concentration curve (Fig. 4).
P1a
4.44(ꢃ0.01)
0.22(ꢃ0.01),
5.1(ꢃ0.1)
520, 0.9992 (28)
With Ka << K0h for both pigments, it can be concluded that the
quinonoid bases (kinetic products) are present in very minor
concentrations at equilibrium (<10% of the total pigment concen-
tration) compared to the chalcone (thermodynamic product) (Fig. 5).
P2 appears significantly less acidic than P1 as a consequence of the
P1b
P2a
3.45(ꢃ0.03)
4.72(ꢃ0.02)
2.09(ꢃ0.04), 0
0.085(ꢃ0.002),
5.64(ꢃ0.15)
e
470, 0.996 (25)
526, 0.998 (26)
P2b
n.a.c
e
a
Spectral measurements immediately after addition of pigment to buffer:
Kobs ¼ Ka. Curve-fitting according to eq. (1).
replacement of the acidic proton at C7-OH by the b-D-glucosyl moiety.
Unexpectedly, P2 turned out to be rather unstable in mildly acidic
conditions (pH > 4e5), thereby complicating the determination of
b
Spectral
measurements
on
solutions
equilibrated
overnight:
Kobs ¼ K0h þ Ka z K0h. Curve-fitting according to eq. (2).
c
Not applicable because of insufficient chemical stability (see section 3.2).

12
N. Mora-Soumille et al. / Dyes and Pigments 96 (2013) 7e15
0.5
0.4
0.3
0.2
0.1
0.0
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
2
3
4
5
6
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Fig. 4. Dependence of the apparent rate constant of flavylium P1 hydration as
a function of the proton concentration. The solid line is the result of the curve-fitting
0.6
0.5
0.4
0.3
0.2
0.1
0.0
according to eq. (5): k0h ¼ 15.4(ꢃ0.4) ꢀ 106 Mꢂ1 sꢂ1, k0
(r ¼ 0.999).
¼ 9.3(ꢃ2.0) ꢀ 10ꢂ5 sꢂ1
ꢂh
A(467 nm)
3.3. Aluminum complexation
Anthocyanins having a 1,2-dihydroxy substitution in their B-
ring can bind hard metal ions such as Al3þ and Fe3þ with
concomitant removal of the phenolic protons [7,12]. The corre-
sponding chelates display a quinonoid structure that is much more
stable than in the absence of metal ions. Consequently, metal
binding is a powerful mechanism of color variation and stabiliza-
tion in plants. In this work, P1 and P2 were compared for their
ability to bind Al3þ in mildly acidic aqueous solutions mimicking
their natural medium (vacuoles of plant cells).
A(367 nm)
0
2
4
6
8
When P1 and P2 are added to a Al3þ solution in a pH 4 acetate
buffer, a relatively fast binding of the colored form is observed
which is manifested by a decay of the visible absorption band at ca.
470 nm (free ligand L, a mixture of flavylium and quinonoid bases)
and a building-up of a broader absorption band at ca. 525 nm
characteristic of the metal complex (ML) (Fig. 6). Interestingly, the
Al3þ complex of P1 appeared much less stable than the one of P2 as
judged from the subsequent decay of A(525 nm) and concomitant
increase of the absorbance at 370e380 nm featuring free chalcone
accumulation (Fig. 7). However, after overnight equilibrium, accu-
mulation of free chalcone was observed with both P1 and P2
although in much lower concentration with the glucoside (Fig. 6).
The absorbanceetime curves for free ligand consumption and
complex and chalcone formation could be analyzed simultaneously
according to a simplified mechanism assuming irreversible metal
binding (rate constant kM) and water addition on both the free
ligand and its complex (respective rate constants kohbs and khM). In
addition to the Beer’s law for the different pigment species (L, ML,
C), eqs. (6)e(9) were used in the curve-fittings:
Fig. 3. Top: pH-dependence of the visible absorbance of P2. Spectral measurements
immediately after addition of pigment to buffer (the solid line is the result of the
curve-fitting according to eq. (1)). Down: Time-dependence of the UVevisible absor-
bance of P2 at pH 4.0 (the solid lines are the result of exponential curve-fittings).
glucosyl moiety. Surprisingly, this is not so with the corresponding
kinetics. Indeed, at pH 4.0, glycosidation at C7-OH lowers the rate
constant of flavylium consumption by a factor ca. 8 (Table 2). As the
observed kinetics is governed by the step of chalcone (Z,E) isomeri-
zation, it may be suggested that the þM effect of O7 weakens the C3]
C4 double bond by enhancing the conjugation with C4]O and that
thiseffectisstronglydecreasedbyglycosidationofO7, therebymaking
the chalcone (Z,E) isomerization much slower for P2 than for P1 with
no impact on the thermodynamics of the reaction.
Â
ÃÂ Ã
 Ã
ꢂd½Lꢄ=dt ¼ kM
M
L þ kohbs
L
(6)
(7)
(8)
(9)
Table 2
Kinetics of flavylium e chalcone conversion at pH 4.0.
Â
ÃÂ Ã
Â
Ã
ꢂd½Mꢄ=dt ¼ kM
M
L ꢂ kMh ML
105kohbs (sꢂ1
)
A0,
D
A
pK0h
l (nm), r
a
Pigment
P1
121.3(ꢃ0.1)
134.1(ꢃ0.3)
16.5(ꢃ0.2)
15.7(ꢃ0.2)
1.70, 1.28
3.38
470, 0.99999
377, 0.99995
467, 0.9998
367, 0.9998
Â
ÃÂ Ã
Â
Ã
L ꢂ kMh ML
0.225, 0.693
0.533, 0.405
0.072, 0.295
d½MLꢄ=dt ¼ kM
M
P2
3.42
 Ã
Â
Ã
d½Cꢄ=dt ¼ kohbs L þ khM ML
a
Calculated according to eq. (3).

N. Mora-Soumille et al. / Dyes and Pigments 96 (2013) 7e15
13
Mole fraction
A
1.4
1.0
1
AH
1.2
C
0.8
0.6
0.4
0.2
1.0
0.8
3
2
0.6
0.4
A
0.2
0.0
0.0
2
300
400
500
λ (nm)
600
700
3
4
pH
5
6
A
Mole fraction
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1.0
1
2
AH
C
0.8
0.6
0.4
0.2
3
A
0.0
2
300
400
500
λ (nm)
600
700
3
4
5
6
pH
Fig. 6. UVevisible spectra of P1 (Top) and P2 (Down) in the presence of Al3þ (4 equiv.)
at pH 4. 1: free ligand, 2: time of maximal Al3þ complexation, 3: after equilibration
overnight.
Fig. 5. Speciation diagrams of P1 (Top) and P2 (Down) at equilibrium.
The maximal absorbance amplitude at 525 nm (DA) was also
The optimized values of the observed rate constants are
collected in Table 3. It is noteworthy that no satisfactory curve-
fitting could be achieved without the hypothesis of water addi-
tion onto the complex. In particular, the slow decay of the
absorption band of the Al3þ e P1 complex could not be accounted
for by assuming reversible metal binding and water addition on the
free ligand only. From this kinetic analysis, it is also clear that Al3þ
binding is faster with P1 than with P2 but that the Al3þ e P1
complex is much more prone to water addition than the Al3þ e P2
complex. From this viewpoint, glycosidation at C7-OH can be
regarded as an efficient way to increase the stability of the Al3þ
complex in mildly acidic aqueous solution. Surprisingly, the values
of the observed rate constant for free flavylium hydration (kohbs) are
higher than those estimated in the absence of Al3þ (Table 2),
especially for P2. It is unclear why the free flavylium e chalcone
conversion is faster in the presence of Al3þ. It can however be noted
that CE itself probably binds Al3þ as evidenced by the weak bath-
ochomic shift (ca. 10 nm) observed when comparing Fig. 1 (lower
part, free CE) and Fig. 6 (lower part). As the kinetics of the overall
flavylium e chalcone conversion is governed by the slow step of
plotted as a function of the total metal concentration (Mt) and the
corresponding curve analyzed by assuming pure 1:1 binding and
negligible chalcone formation (eqs. (10) and (11), KM: metal e
pigment binding constant, Lt: total ligand concentration,
D
3
¼
3
525 ꢂ
3
525).
L
ML
D
A ¼
D
3
ðMt ꢂ ½MꢄÞ
(10)
(11)
ꢂ
ꢃ
Â
Ã
KMLt
1 þ KM½Mꢄ
Mt
¼
M
1 þ
The following KM values were thus obtained: for P1,
KM
¼
8.9(ꢃ1.3)
ꢀ
103 Mꢂ1 (r
¼
0.997), for P2,
KM ¼ 19(ꢃ8) ꢀ 103 Mꢂ1 (r ¼ 0.98).
Alternatively, for P2, the observed first-order rate constant of
metal binding (koMbs) was found to vary linearly with the total metal
concentration (Fig. 8, r ¼ 0.998). From koMbs ¼ kMMt þ kꢂM, estimates
for the rate constants of complex formation and dissociation were
obtained: kM ¼ 44(ꢃ2) Mꢂ1
s
ꢂ1, k
¼ 20(ꢃ4) ꢀ 10ꢂ4
s
ꢂ1, from
ꢂM
which one deduces: KM ¼ kM/k
¼ 22 ꢀ 103 Mꢂ1
.
ꢂM
(Z)-(E) chalcone isomerization, it can be suggested that Al3þ
e
Finally, for P1, the overall first-order rate constant of flavylium e
chalcone conversion was plotted as a function of Mt and the
chalcone binding accelerates the latter step.

14
N. Mora-Soumille et al. / Dyes and Pigments 96 (2013) 7e15
1.2
1.0
0.8
0.6
0.4
0.2
0.0
25
20
15
10
A(474 nm)
A(525 nm)
A(377 nm)
5
0
0
50
100
150
Time (s)
200
250
300
0.0
0.1
0.2
0.3
0.4
0.5
0.6
0.5
0.4
0.3
0.2
0.1
0.0
Fig. 8. Plot of the apparent first-order rate constant of Al3þ e P2 binding as a function
of the total metal concentration (pH 4.0).
A(525 nm)
Thus, one obtains estimates for the binding constant as well as
for the rate constants of chalcone formation from the free flavylium
and
its
complex:
kohbs
¼
195(ꢃ12)
ꢀ
10ꢂ5
sꢂ1
,
kMh ¼ 45(ꢃ1) ꢀ 10ꢂ5 sꢂ1, KM ¼ 9.8(ꢃ3.5) ꢀ 103 Mꢂ1
.
A(470 nm)
In summary, P1 and P2 display similar affinities for Al3þ
(KM ¼ 1e2 ꢀ 104
M
ꢂ1) but the complex of P2 is more stable as its
A(370 nm)
conversion into the corresponding chalcone is slower.
3.4. Scavenging of the DPPH radical
0
100
200
300
400
500
600
Polyphenols with a catechol group are potent electron or H-
atom donors to radicals due to the relatively stability of the semi-
quinone radicals thus formed (a combination of electronic and
intramolecular hydrogen bonding effects). Semiquinone radicals
derived from polyphenols are however rapidly converted into o-
quinones by disproportionation or scavenging of a second radical.
From these reactive intermediates, different pathways (oligomeri-
zation, solvent addition, ring cleavage) typically ensue, ultimately
Time (s)
Fig. 7. Time-dependence of the UVevisible absorbance of P1 (Top) and P2 (Down) after
addition of Al3þ (4 equiv.) at pH 4.0. The solid lines are the result of the curve-fittings
according to eqs. (6)e(9).
corresponding curve analyzed according to eqs. (11) and (12) (Fig. 9,
r ¼ 0.996).
Â
Ã
kohbs þ kMh KM
M
kohverall
¼
(12)
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
1 þ KM½Mꢄ
Table 3
Kinetic analysis of Al3þ binding and water addition at pH 4.0a.
Pigment Al3þ
(equiv.) (Mꢂ1 sꢂ1
kM
105kohbs, 105khM (sꢂ1
)
3
( 3 )
c) (Mꢂ1 cmꢂ1
b
ML C
)
P1
P2
2
4
6
8
10
2
4
6
8
10
486(ꢃ3)
288(ꢃ1)
257(ꢃ2)
195(ꢃ1)
161(ꢃ1)
302(ꢃ7), 82(ꢃ1)
324(ꢃ6), 64(ꢃ1)
341(ꢃ14), 54(ꢃ1)
614(ꢃ23), 38(ꢃ1)
467(ꢃ26), 29(ꢃ1)
19990, 27980 (21920)
21290, 21850 (20480)
24690, 20550 (21960)
23000, 17400 (19950)
27550, 18840 (22390)
25520, 20600 (4340)
23.2(ꢃ1.0) 284(ꢃ9), 6.3(ꢃ0.2)
37.3(ꢃ1.3) 226(ꢃ22), 3.2(ꢃ0.3) 16240, 10420 (8140)
41.3(ꢃ1.4) 284(ꢃ36), 2.5(ꢃ0.2)
40.3(ꢃ1.1) 315(ꢃ36), 1.8(ꢃ0.2)
12.6(ꢃ0.1) 182(ꢃ6), 3.7(ꢃ0.1)
16550, 9200 (10440)
16430, 8970 (11990)
15890, 10020 (9250)
0.0
0.1
0.2
0.3
0.4
0.5
a
P1: simultaneous curve-fitting of the absorbance curves at 525, 474 and 377 nm
t ¼ 2 s, tf ¼ 5 min) according to eqs. (6)e(9). P2: simultaneous curve-fitting of the
(D
absorbance curves at 525, 470 and 370 nm (
D
t ¼ 10 s, tf ¼ 15 min).
Fig. 9. Plot of the apparent first-order rate constant of flavylium e chalcone conversion
for P1 as a function of the total metal concentration (pH 4.0). The solid line is the result
of the curve-fitting according to eq. (12).
b
c
First value at 525 nm, second value at 474 nm (P1) or 470 nm (P2).
At 377 nm (P1) or 370 nm (P2).

N. Mora-Soumille et al. / Dyes and Pigments 96 (2013) 7e15
15
0
0
,
giving a complex set of oxidation products [4]. Such reactions may
take place in plants where polyphenols act as electron donors for
the enzymatic reduction of hydrogen peroxide in vacuoles but also
in food and in the digestive tract where polyphenols may protect
polyunsaturated lipids from oxidation and/or regenerate the potent
from C4 -OH. Alternatively, the anionic quinonoid base A4 ,7
although in minor concentration, could be the true electron-donor
due to the extended electron-delocalization on the chromophore.
Indeed, upon deprotonation, A4 and A7 merge into a single
0
p-
electron-rich anion with a concomitant spectacular rise of the
HOMO by ca. 4 eV (HyperChem software, AM1 parametrization).
On the other hand, the total stoichiometry value is defined by
the fate of the antioxidant during oxidative degradation. In
a previous work [15], we showed that upon two-electron oxidation
P1 forms an o-quinone that undergoes water addition on the C- and
A-rings. The quinonoid bases thus formed display an additional OH
groups and must also participate in radical-scavenging, thereby
prolonging the activity. It is possible that glycosidation at C7-OH
somehow hinders these sequences of oxidation/water addition,
thus resulting in lower ntot values.
phenolic antioxidant a-tocopherol (vitamin E) by H-atom or elec-
tron transfer to its aryloxyl radical. Whether in plants, food or the
digestive tract, anthocyanins typically experience a mildly acidic
aqueous environment. Hence, investigating their electron-donating
capacity toward the stable radical DPPH (1,1-diphenyl-2-
picrylhydrazyl) in a MeOH/pH 3.5 acetate buffer (1:1) mixture
(final pH ca. 4.4) is a first acceptable approach to assess their
antioxidant activity.
When the pigment is added to the DPPH solution, a fast decrease
of the DPPH visible absorption band is observed featuring the
transfer of the labile H-atoms of the catechol OH groups. Then,
a slower decay follows that reflects the residual H-donating activity
of some oxidation products. From the total absorbance amplitude
4. Conclusion
over 1 h (
D
A515 ¼ A515(t ¼ 1 h) ꢂ A515(t ¼ 0, i.e. before pigment
30,40-Dihydroxy-7-O-
b-D-glucopyranosyloxyflavylium chloride
addition)), the total number of DPPH radicals reduced to the corre-
sponding hydrazine per pigment molecule (total stoichiometry ntot
)
(P2), which has been synthesized in this work at a scale of several
hundreds of mg, is a valuable model of natural anthocyanins. Its
glucose moiety improves the water solubility while only moder-
ately affecting the acidebase and hydration properties of the
chromophore. Due to its catechol group, P2 can be used to inves-
tigate the affinity of anthocyanins for metal ions and their ability to
deliver electrons to radicals, two characteristics that have a strong
impact on the pigmentation and nutritional properties of
anthocyanins.
can be deduced from the following relationship: ntot
¼
D /
A515
515
515
(
3
DPPH [pigment]0) (Table 4) with
3
¼ 11240 Mꢂ1 cmꢂ1 [13].
DPPH
The fast step was kinetically analyzed according to a model
developed in details in our previous work [13,14] and permitting
the estimation of the rate constant of first H-atom transfer (k1) and
the partial stoichiometry n (number of DPPH radicals reduced per
pigment molecule during the fast step) (Table 4). It is noteworthy
that the kinetic analysis had to be conducted at 600 nm so as to
avoid interference with the visible absorption of the pigment at the
onset of the reaction.
As the catechol group is the critical determinant of the H-
donating activity of polyphenols, it may be anticipated that P2 be as
good at scavenging the DPPH radical as P1. Surprisingly, this is not
so. Not only does P1 transfer a first H-atom to DPPH ca. 4 times as
rapidly as P2, but also the total number of DPPH radicals reduced by
P1 is ca. 3 times as high as by P2. As DPPH is a rather bulky radical,
part of the lower reactivity of P2 could be attributed to the steric
hindrance brought about by the glucose moiety. It is also important
to keep in mind that P2 is less acidic than P1. As the flavylium ion
(because of its positive charge) is expected to be a poor electron or
H-atom donor and as the kinetic of flavylium e chalcone conver-
sion is much slower than the DPPH-scavenging reaction, rate
constant k1 must essentially reflect the ability of the quinonoid
bases to transfer an H-atom to DPPH. Moreover, P1 can form both
References
[1] Quideau S, Deffieux D, Douat-Casassus C, Pouysegu L. Plant polyphenols:
chemical properties, biological activities, and synthesis. Angew Chem Int Ed
2011;50:586e621.
[2] Perez-Jimenez J, Neveu V, Vos F, Scalbert A. Systematic analysis of the content
of 502 polyphenols in 452 foods and beverages: an application of the phenol-
explorer database. J Agric Food Chem 2010;58:4959e69.
[3] Crozier A, Jaganath IB, Clifford MN. Dietary phenolics: chemistry, bioavail-
ability and effects on health. Nat Prod Rep 2009;26:1001e43.
[4] Dangles O. Antioxidant activity of plant phenols: chemical mechanisms and
biological significance. Curr Org Chem 2012;16:692e714.
[5] Frankel EN, Finley JW. How to standardize the multiplicity of methods to
evaluate natural antioxidants. J Agric Food Chem 2008;56:4901e8.
[6] Pina F, Melo MJ, Laia CAT, Parola AJ, Lima JC. Chemistry and applications of
flavylium compounds: a handful of colours. Chem Soc Rev 2012;41:869e908.
[7] Yoshida K, Mori M, Kondo T. Blue flower color development by anthocyanins:
from chemical structure to cell physiology. Nat Prod Rep 2009;26:857e964.
[8] Tsuda T. Dietary anthocyanin-rich plants: biochemical basis and recent
progress in health benefits studies. Mol Nutr Food Res 2012;56:159e70.
0
0
A4 and A7 tautomers whereas A4 is the only possible quinonoid
base for P2. Overall, it could be suggested that A7 makes a major
contribution to the radical-scavenging activity of P1 via H-transfer
[9] El Hajji H, Dangles O, Figueiredo P, Brouillard R. 30-(
b-D-Glycopyranosyloxy)
flavylium ions: synthesis and investigation of their properties in aqueous
solution. Hydrogen bonding as a mean of colour variation. Helv Chim Acta
1997;80:398e413.
[10] Dangles O, Saito N, Brouillard R. Kinetic and thermodynamic control of fla-
vylium hydration in the pelargonidinecinnamic acid complexation. Origin of
the extraordinary flower color diversity of Pharbitis nil. J Am Chem Soc 1993;
115:3125e32.
[11] Nave F, Petrov V, Pina F, Teixeira N, Mateus N, de Freitas V. Thermodynamic
and kinetic properties of a red wine pigment: catechin-(4,8)-malvidin-3-O-
glucoside. J Phys Chem B 2010;114:13487e96.
[12] Dangles O, Elhabiri M, Brouillard R. Kinetic and thermodynamic investigation
of the aluminiumeanthocyanin complexation in aqueous solution. J Chem Soc
Perkin Trans 1994;2:2587e96.
[13] Goupy P, Loonis M, Dufour C, Dangles O. A quantitative kinetic analysis of
hydrogen transfer reactions from dietary polyphenols to the DPPH radical.
J Agric Food Chem 2003;51:615e22.
[14] Goupy P, Bautista-Ortin AB, Fulcrand H, Dangles O. Antioxidant activity of
wine pigments derived from anthocyanins: hydrogen transfer reactions to the
DPPH radical and inhibition of the heme-induced peroxidation of linoleic acid.
J Agric Food Chem 2009;57:5762e70.
[15] Dangles O, Fargeix G, Dufour C. Antioxidant properties of anthocyanins and
tannins: a mechanistic investigation with catechin and the 30,40,7-trihydrox-
yflavylium ion. J Chem Soc Perkin Trans 2000;2:1653e63.
Table 4
DPPH scavenging by P1 and P2 in MeOH/pH 3.5 acetate buffer (1:1).
a
Pigment, conc. (
m
M)
DPPH conc. (
m
M)
k1 (Mꢂ1 sꢂ1
4305(ꢃ64)
)
na
P1, 10
P1, 15
P1, 15
P2, 15
P2, 15
P2, 25
P2, 25
P2, 30
77.0
83.8
77.4
64.6
64.8
70.8
66.5
65.9
2.22(ꢃ0.01)
2.09(ꢃ0.03)
2.67(ꢃ0.01)
0.85(ꢃ0.01)
0.70(ꢃ0.01)
0.91(ꢃ0.01)
0.93(ꢃ0.01)
0.82(ꢃ0.01)
4290(ꢃ160)
5354(ꢃ70)
1426(ꢃ37)
1355(ꢃ48)
1368(ꢃ32)
1363(ꢃ11)
1066(ꢃ10)
From the absorbance amplitude at 515 nm over 1 h: total number of DPPH radicals
reduced per pigment molecule ntot ¼ 4.0(ꢃ0.2) for P1, 1.38(ꢃ0.04) for P2 (means
from triplicates).
a
Curve-fitting of A600 vs. time curves (see model in text) over the 40 first seconds
following pigment addition: k1: rate constant of first H-atom abstraction, n: number
of DPPH radicals reduced per pigment molecule.