Kinetic study of the quenching reaction of singlet oxygen by flavonoids in ethanol solution

Article abstract of DOI:10.1021/jp0451389

The quenching rate of singlet oxygen (¹O₂) by seven kinds of flavonoids (flavone, flavonol, chrysin, apigenin, rutin, quercetin, and myricetin) with 2,3-double bonds has been measured spectrophotometrically in ethanol at 35°C. The overall rate constants k_Q (=k_q + k_t, physical quenching + chemical reaction) increased as the number of OH groups substituted to the flavone skeleton (that is, the total electron-donating capacity of flavonoids) increases. The existence of catechol or pyrogallol structure in the B-ring is essential for the ¹O ₂ quenching of flavonoids. Log k_Q was found to correlate with their peak oxidation potentials, E_P; the flavonoids that have smaller E_P values show higher reactivities. Similarly, log k _Q values of flavonoids correlate with the energy level of the highest occupied molecular orbital (E_HOMO), calculated by the PM3 MO method, and the longest wavelength ππ* excitation energy (E_ex). The contribution of the chemical reaction (k_r) was found to be negligible in these flavonoids. The k_Q values of rutin, quercetin, and myricetin [(1.21-5.12) × 10⁸ M^-1 s^-1] were found to be larger than those of lipids [(0.9-6.4) × 10⁴ M^-1 s^-1], amino acids (<3.7 × 10⁷ M^-1 s^-1), and DNA (5.1 × 10⁵ M ^-1s^-1). The result suggests that these flavonoids may contribute to the protection of oxidative damage in foods and plants, by quenching ¹O₂.

Full text of DOI:10.1021/jp0451389

4234
J. Phys. Chem. B 2005, 109, 4234-4240
Kinetic Study of the Quenching Reaction of Singlet Oxygen by Flavonoids in Ethanol
Solution
Souichi Nagai, Keishi Ohara, and Kazuo Mukai*
Department of Chemistry, Faculty of Science, Ehime UniVersity, Matsuyama 790-8577, Japan
ReceiVed: October 25, 2004; In Final Form: December 27, 2004
The quenching rate of singlet oxygen (1O2) by seven kinds of flavonoids (flavone, flavonol, chrysin, apigenin,
rutin, quercetin, and myricetin) with 2,3-double bonds has been measured spectrophotometrically in ethanol
at 35 °C. The overall rate constants kQ ()kq + kr, physical quenching + chemical reaction) increased as the
number of OH groups substituted to the flavone skeleton (that is, the total electron-donating capacity of
1
flavonoids) increases. The existence of catechol or pyrogallol structure in the B-ring is essential for the O2
quenching of flavonoids. Log kQ was found to correlate with their peak oxidation potentials, EP; the flavonoids
that have smaller EP values show higher reactivities. Similarly, log kQ values of flavonoids correlate with the
energy level of the highest occupied molecular orbital (EHOMO), calculated by the PM3 MO method, and the
longest wavelength ππ* excitation energy (Eex). The contribution of the chemical reaction (kr) was found to
be negligible in these flavonoids. The kQ values of rutin, quercetin, and myricetin [(1.21∼5.12) × 108 M-1
s-1] were found to be larger than those of lipids [(0.9∼6.4) × 104 M-1 s-1], amino acids (<3.7 × 107 M-1
s-1), and DNA (5.1 × 105 M-1 s-1). The result suggests that these flavonoids may contribute to the protection
1
of oxidative damage in foods and plants, by quenching O2.
1. Introduction
flavonoids increase in the order
Flavonoids are natural polyphenolic compounds widely
distributed in foods and plants. Flavonoids display pronounced
biological activities, protecting against coronary heart disease
(CHD),1,2 cancer,3-5 inflammation,5 etc. The biological activities
of flavonoids have been related to their properties as antioxi-
dants, and many studies have been performed on the inhibition
of lipid peroxidation in biological systems or model media
[microsomes, mitochondria, liposomes, low-density lipoprotein
(LDL), etc.].6-10 A notable feature of flavonoids is their high
reactivity toward active free radical species. Several kinetic
studies have been performed on the reaction of flavonoids with
active free radicals (N3•, HO•, O2- •, t-BuO•, and LOO•) by
use of pulse radiolysis techniques11-14 and high-performance
liquid chromatography (HPLC).15,16
In a previous work, to clarify the structure-activity relation-
ship in the scavenging reaction of free radical by flavonoids,
we have measured the second-order rate constants (kS and kR)
for the reaction of six kinds of flavonoids (flavone, chrysin,
flavonol, apigenin, rutin, and quercetin) with 2,6-di-tert-butyl-
4-(4-methoxyphenyl)phenoxyl (ArO•, abbreviated to aroxyl) and
5,7-diisopropyltocopheroxyl (Toc•) radicals in ethanol, 2-pro-
panol/water (5:1 v/v), and aqueous Triton X-100 micellar
solutions (5.0 wt %) (reactions 1 and 2):17
flavone < chrysin < flavonol < apigenin < rutin <
quercetin (3)
independent of both pH value and solvent system. Rutin and
quercetin, with 3′- and 4′-OH groups at the B-ring, showed high
reactivity, indicating that the catechol structure in the B-ring is
the obvious radical target site for flavonoids.11,12 It was found
that quercetin and rutin have high activity in vitamin E
regeneration.17
Illumination with excess photosynthetically active radiation
(PAR) is a stress factor for plants, known as photoinhibition. It
has been found that photoinhibition of photosynthesis in broad
bean leaves is accompanied by singlet oxygen (1O2) produc-
1
tion.18 Recently, it has been reported that O2 is generated in
the UVA-irradiated skin of live mice.19 Flavonoids are widely
present in foods and plants in high concentration and may
1
function as quenchers of O2 in biological systems. Conse-
quently, measurements of the quenching rate of 1O2 with
flavonoids are very important. However, the examples are very
limited, as described below.
The quenching rates kQ () kq + kr, physical quenching +
chemical reaction) of singlet oxygen with catechins [(+)-
catechin (CA), epicatechin (EC), epigallocatechin (EGC), epi-
catechingallate (ECG), and epigallocatechingallate (EGCG)] in
CH3CN have been reported by Jovanovic et al.,14 by a laser
flash photolysis method:
kS
ArO• + flavonoid
Toc• + flavonoid
9
8 ArOH + flavonoid•
(1)
(2)
kR
kQ
98
1O2 + flavonoid
98 TocH + flavonoid•
physical quenching (kq) + chemical reaction (kr) (4)
The rate constants (kS and kR) obtained in micellar solution
showed notable pH dependence. The kS and kR values of
where the former results in energy transfer and de-excitation
of the singlet state but no chemical change in the energy
acceptor. The latter results in modification of the target. The
* Corresponding author: tel 81-89-927-9588; fax 81-89-927-9590; e-mail
mukai@chem.sci.ehime-u.ac.jp.
10.1021/jp0451389 CCC: $30.25 © 2005 American Chemical Society
Published on Web 02/10/2005

Quenching of Singlet Oxygen by Flavonoids in Ethanol
J. Phys. Chem. B, Vol. 109, No. 9, 2005 4235
numbers (n ) 0∼6) of OH substituents on a flavone skeleton,
have been measured in ethanol at 35 °C, by a competition
reaction method21-23 (see Schemes 1 and 2). The rate constants
of two structurally related flavonoids (naringenin and taxifolin)
without a 2,3-double bond were measured to investigate the
effect of π-conjugation between A- (resorcinol) and B- (phenol,
catechol and pyrogallol) rings on the reaction rate. The chemical
1
reaction (kr) of flavonoids with O2 has been studied spectro-
photometrically, by reacting flavonoids with 1O2. PM3 molecular
orbital (MO) calculations and the measurements of UV-vis
absorption spectra were performed for these flavonoids. From
the results, the structure-activity relationship in the 1O2
quenching reaction of flavonoids has been discussed.
2. Experimental Section
2.1. Materials. All the flavonoids used in the present work
are commercially available: flavone (Kanto Chemicals), fla-
vonol (Wako Chemicals, Japan), chrysin (Aldrich), apigenin
(Tokyo Kasei Organic Chemicals, Japan), rutin (Nakarai
Chemicals, Japan), quercetin (Aldrich), myricetin (Aldrich),
naringenin (Aldrich), and taxifolin (Aldrich). 3-(1,4-Epidioxy-
4-methyl-1,4-dihydro-1-naphthyl)propionic acid (endoperoxide,
EP) (see Scheme 1) was prepared by the published procedure.21
2,5-Diphenyl-3,4-benzofuran [DPBF (Tokyo Kasei Chemicals,
Japan) (see Figure 1)] is commercially available.
2.2. Measurements. The measurements of rate constant were
performed on a Shimadzu UV-2100S spectrophotometer. All
measurements were performed at 35.0 ( 0.5 °C.
2.3. Calculations. The semiempirical MO calculations were
performed with MOPAC 2000 ver. 1.0 on Windows XP with
parameters in ref 24. The optimized geometries, energy param-
eters, and MO coefficients were calculated by the semiempirical
PM3 method. Molecular geometries were totally optimized.
3. Results
1
3.1. Overall Rate Constants (kQ) for the Reaction of O2
with Flavonoids. Singlet oxygen was generated by the thermal
decomposition of the endoperoxide (EP).21,22 2,5-Diphenyl-3,4-
benzofuran (DPBF) was used as standard compound. The overall
rate constants kQ ()kq + kr) for the reaction of 1O2 with
flavonoids were determined in ethanol by eq 5 derived from
the steady-state treatment of Scheme 2:23
Figure 1. Molecular structures of flavone (Fl), flavonol, chrysin (Ch),
apigenin (Ap), rutin (Ru), quercetin (Qu), myricetin (My), naringenin
(Na), taxifolin (Ta), and DPBF and numbering system.
notable differences in the rate constants (kQ) have not been
observed for catechins; the values obtained are 1.1 × 108, 9.6
× 107, 1.1 × 108, 2.2 × 108, and 2.2 × 108 M-1 s-1 for CA,
EC, EGC, ECG, and EGCG, respectively. The difference in the
rate constants is less than 3-fold. A kinetic study of the
quenching reaction of 1O2 by 13 kinds of flavonoids (from the
flavonol, flavone, flavanone, and flavane families) in CH3OH
S0/SS ) 1 + [(kq + kr)/kd][flavonoid]
(5)
where S0 and SS are slopes of the first-order plots of disappear-
1
ance of O2 acceptor, DPBF, in the absence and presence of
1
flavonoid, respectively. kd is the rate of deactivation of O2 in
1
has been performed by Tournaire et al.,20 using a near-IR O2
ethanol.
luminescence method. They reported that the efficiency of the
physical quenching (kq) is mainly controlled by the presence of
a catechol moiety on ring B, whereas the structure of ring C
(particularly the presence of a hydroxyl group activating the
2,3-double bond) is the main factor determining the efficiency
of the chemical reactivity (kr) of flavonoids with 1O2. The total
reactivity scale (kQ) is dominated by kq, which is in general
higher than kr. (+)-Catechin showed the highest overall quench-
ing rate (kQ ) kq + kr ) 5.8 × 106 M-1 s-1 in CH3OH) among
the above flavonoids. However, the kQ value obtained for (+)-
catechin is about 20 times smaller than that (1.1 × 108 M-1
s-1) reported by Jovanovic et al.14
Figure 2 shows an example of the interaction between DPBF
(5.98 × 10-5 M) and EP (2.94 × 10-4 M) without flavonoids
in ethanol solution at 35 °C. By the reaction, the appearance of
absorption of EP precursor at λmax ) 288 nm due to the thermal
decomposition of EP (and, thus, the production of 1O2) and the
disappearance of DPBF at λmax ) 411 nm due to the chemical
reaction between DPBF and 1O2 produced were observed
simultaneously at 288 and 411 nm, respectively (see Table 1).
The pseudo-first-order rate constant (S0) was obtained by
following the decrease in absorbance at 411 nm of the DPBF.
Similarly, solutions containing EP (3.69 × 10-4 M), DPBF
(7.02 × 10-5 M), and various amounts of rutin (0∼1.17 × 10-3
M) in ethanol were reacted at 35 °C. The disappearance of
DPBF was measured at 411 nm.22,23 A plot of S0/SS vs
concentration of rutin is shown in Figure 3. The overall rate
In the present work, the quenching rates kQ of 1O2 by seven
kinds of flavone derivatives (flavone, flavonol, chrysin, apigenin,
rutin, quercetin, and myricetin; see Figure 1), having different

4236 J. Phys. Chem. B, Vol. 109, No. 9, 2005
Nagai et al.
SCHEME 1
TABLE 1: UV-Vis Absorption Maxima (λmax) and Molar
Extinction Coefficients (E) of the Flavonoids and Related
Compounds in Ethanol, and Energy Level of HOMO
SCHEME 2
(EHOMO
)
ꢀ,
flavonoids
flavone
λ
max, nm
L mol-1 cm-1
EHOMO, eV
-9.29
294
251
343
305
240
313
269
246 (sh)
336
289 (sh)
269
361
295 (sh)
259
371
301 (sh)
256
377
305
255
23 300
19 400
19 000
13 600
21 300
17 600
43 600
21 200
19 800
12 300
17 300
17 500
7610
20 300
20 300
6440
19 000
59 200
18 200
50 000
52 000
51 600
6510
flavonol
chrysin
apigenin
rutin
-8.97
-9.24
-9.16
-9.06
-9.05
-9.06
constants (kQ) were calculated by using the value of kd in ethanol
(kd ) 8.3 × 104 s-1), reported by Merkel and Kearns.25
1
Similarly, flavonoids were reacted with O2 in ethanol. S0/SS
vs [flavonoid] plots for apigenin and myricetin are also shown
in Figure 3. The kQ values obtained were summarized in Table
2, together with those reported for R- and γ-tocopherol,26-29
ubiquinol-10,30 and γ-tocopherol hydroquinone (plastoquinol
model).30 The experimental error in kQ value for each flavonoid
was (8% at maximum.
1
3.2. Chemical Reaction of O2 with Flavonoids. Endoper-
quercetin
myricetin
oxide (EP) was prepared by the reaction of 3-(4-methyl-1-
1
naphthyl)propionic acid (EP precursor) with O2 produced by
the photosensitization reaction of methylene blue (MB) in
ethanol (see Scheme 1).21 The NMR measurement indicates that
the powder sample of EP includes about 19% unreacted EP
precursor. The UV absorption spectra of (a) EP (including about
24% EP precursor) and (b) EP precursor in ethanol at 25 °C
are shown in Figure 4, where both [EP] and [EP precursor] are
1.50 × 10-4 M. EP precursor shows absorption maxima at λmax
(ꢀ) ) 299 (sh) (4470), 288 (6560), 279 (5340), and 233 (9430),
as listed in Table 1. In fact, the UV absorption spectrum of EP
in ethanol shows a maximum at λmax ) 288 nm due to about
24% EP precursor in addition to a maximum at λmax ) 229 nm
(ꢀ ) 6510) due to intrinsic EP. Rutin shows two absorption
maxima [λmax (ꢀ) ) 361 nm (17 500) and 259 nm (20 300)] at
different wavelength regions (see Table 1).
naringenin
taxifolin
EP
290
291
229
-9.27
-9.18
EP precursor
299 (sh)
288
279
233
411
311
262
4470
6560
5340
9430
19 700
7260
25 900
DPBF
Figure 3. Plot of S0/SS vs concentrations of apigenin, rutin, and
myricetin.
Figure 2. Change in electronic absorption spectrum of DPBF and
endoperoxide (EP) during reaction of DPBF with EP in ethanol solution
at 35 °C. [DPBF]t)0 ) 5.98 × 10-5 M and [EP]t)0 ) 2.94 × 10-4 M.
The spectra were recorded at 20 min intervals. The arrow indicates
decreasing absorbance of DPBF and increasing absorbance of EP
precursor with time, respectively.
Figure 5 shows an example of the interaction between rutin
(4.50 × 10-5 M) and EP (1.25 × 10-4 M) in ethanol solution
at 35 °C. The spectra were recorded at 30 min intervals. At 35

Quenching of Singlet Oxygen by Flavonoids in Ethanol
J. Phys. Chem. B, Vol. 109, No. 9, 2005 4237
TABLE 2: Second-Order Rate Constants (kQ) and Relative
Rate Constants [100kQ(AO)/kQ(r-Toc)] for the Reaction of
1O2 with Flavonoids and Related Compounds in Ethanol,
and Peak Oxidation Potentials (EP)a
100kQ(AO)/ EP, V vs
antioxidants
flavone
flavonol
chrysin
apigenin
rutin
quercetin
myricetin
naringenin
n
kQ, M-1 s-1 kQ(R-Toc) Ag/AgClb
0 <3 × 105
<0.145
2.59
9.76
13.8
58.7
222
1
2
3
4
5
6
3
5
5.33 × 106
2.01 × 107
2.84 × 107
1.21 × 108
4.57 × 108
5.12 × 108
3.68 × 106
9.37 × 106
2.06 × 108 c
1.38 × 108 c
1.17 × 108 c
0.794
0.658
0.360
0.178
249
1.79
4.54
100
67.0
56.8
0.688
0.248
taxifolin
R-tocopherol
γ-tocopherol
γ-tocopherol
hydroquinone (γ-TQH2)
(plastoquinol model)
ubiquinol-10 (UQ10H2)
â-carotene
a n denotes number of OH groups substituted to flavone skeleton
b Values reported by Hotta et al. (31). c Values reported in a previous
paper (30).
Figure 5. Increase of the absorption of EP precursor at 288 nm due
to the generation of singlet oxygen from endoperoxide (EP) in ethanol
at 35 °C. [EP] ) 1.25 × 10-4 M and [rutin] ) 4.50 × 10-5 M. The
spectra were recorded at 30 min intervals.
1.58 × 108 c
76.7
1.58 × 1010 7670
Figure 6. Plot of log kQ vs EHOMO for flavonol, chrysin, apigenin,
rutin, quercetin, myricetin, naringenin, and taxifolin.
C-ring show chemical reactivity, although kq is in general higher
than kr. On the other hand, the chemical reaction was not
observed for the flavonoids without the 2,3-double bond.
However, the result obtained here shows that the chemical
reaction is almost negligible in all the flavonoids studied. The
reason for such a difference in the chemical reactivity (kr) is
not clear at present.
3.3. Molecular Orbital Calculations of Flavonoids. Mo-
lecular orbital calculations were performed for the flavonoids
by the semiempirical PM3 method.24 The energy levels of the
highest occupied molecular orbital (EHOMO) obtained are listed
in Table 1. The values of log kQ for flavonoids have been plotted
against EHOMO. As shown in Figure 6, log kQ of flavonoids
except for flavonol correlates well with EHOMO with a slope of
8.2 ( 1.6 eV-1 (correlation coefficient ) 0.91). The flavonoids
that have higher EHOMO values show higher reactivities with
singlet oxygen.
Figure 4. UV absorption spectra of endoperoxide (EP) and EP
precursor in ethanol at 35 °C. [EP] ) [EP precursor] ) 1.50 × 10-4
M.
°C, EP decomposes and generates singlet oxygen and EP
precursor. The arrow indicates an increase in absorbance of EP
precursor at 288 nm with time. However, the absorption of rutin
at λmax ) 361 nm shows no change in the intensity. The
1
chemical reaction between O2 and rutin is very slow, and the
measurable change in the absorption spectrum of rutin was not
observed. Similar measurements were performed for the reaction
between EP and flavonoids with and without a 2,3-double bond.
The chemical reaction expected was not observed for all the
flavonoids studied. Consequently, the kQ values obtained for
flavonoids are thought to be due to physical quenching (kq);
that is, kQ ≈ kq.
It has been reported that tocopherols can act as efficient
1
scavengers of O2.26-29 It was shown that R-tocopherol scav-
3.4. UV-Vis Absorption Spectra of Flavonoids. Measure-
ments of UV-vis absorption spectra have been performed for
the flavonoids in ethanol solution. Absorption spectra of
flavonoids with 2,3-double bonds are shown in Figure 7. The
wavelengths of absorption maxima (λmax) and molar absorption
coefficients (ꢀ) obtained are listed in Table 1, indicating that
the values of ꢀ at λmax are similar to each other in these
1
enges O2 by a combination of physical quenching (kq) and
chemical reaction (kr). For instance, the kQ value (kQ ) kq +
kr) of R-tocopherol is 2.5 × 108 M-1 s-1 in pyridine. Because
kq . kr, the quenching process is almost entirely “physical”;
that is, R-tocopherol deactivates about 120 1O2 molecules before
being destroyed by chemical reaction (eq 4).26,27 As described
in the Introduction, Tournaire et al.20 reported that the flavonoids
with a hydroxyl group activating the 2,3-double bond at the
L
flavonoids. The λmax values of the longest wavelength in
flavonoids, except for flavonol, increase with increasing number

4238 J. Phys. Chem. B, Vol. 109, No. 9, 2005
Nagai et al.
indicates that the catechol structure in the B-ring of rutin and
1
quercetin mainly contributes to the quenching of O2. The kQ
value (5.12 × 108 M-1 s-1) obtained for myricetin, which has
3′-, 4′-, and 5′-OH groups at the B-ring, is larger than those of
rutin and quercetin.
1
As described above, the quenching rates of O2 oxygen for
flavonoids increase in the order
flavone (n ) 0) < flavonol (n ) 1) < chrysin (n ) 2) <
apigenin (n ) 3) < rutin (n ) 4) < quercetin (n ) 5) <
myricetin (n ) 6) (6)
in ethanol solution, where n is the number of OH substituents
in these flavonoids. The result clearly indicates that the
quenching rate increases with increasing number of OH sub-
stituents, that is, the electron-donating capacity of these fla-
vonoids. Especially, the existence of catechol or pyrogallol
Figure 7. Absorption spectra of flavone, flavonol, chrysin, apigenin,
rutin, and quercetin. [Flavonoid] ) 1.50 × 10-5 M.
1
structure in the B-ring is essential for the O2 quenching of
flavonoids, although the resorcinol structure in the A-ring also
1
contributes partly (4∼17%) to the O2 quenching.
Apigenin and naringenin have 5- and 7-OH groups at the
A-ring and a 4′-OH group at the B-ring. Further, only the
apigenin has a 2,3-double bond, which is responsible for
π-conjugation between A- and B-rings. In fact, the quenching
rate (kQ) of apigenin (2.84 × 107 M-1 s-1) is 7.7 times larger
than that of naringenin (3.68 × 106 M-1 s-1) without a 2,3-
double bond in ethanol. Similarly, the kQ of quercetin (4.57 ×
108 M-1 s-1) and rutin (1.21 × 108 M-1 s-1) are 49 and 13
times larger than that of taxifolin (9.37 × 106 M-1 s-1) in
ethanol, respectively. The result indicates that the existence of
the 2,3-double bond in quercetin, rutin, and apigenin is important
1
for O2 quenching.
Bors et al.11,12 measured the second-order rate constants for
the reaction of flavonoids with HO•, N3•, and t-BuO• radicals,
at pH 11.5, by a pulse radiolysis technique and reported that
three structural groups in flavonoids are important determinants
for radical scavenging: (i) the 3′- and 4′-OH groups (catechol
structure) in B-ring, which are the obvious radical target site
for all flavonoids; (ii) the 2,3-double bond in conjugation with
a 4-oxo function, which is responsible for electron delocalization
from the B-ring; and (iii) the existence of both 3- and 5-OH
groups for maximal radical-scavenging activity. The results
obtained in the present work indicate that all the above
determinants i-iii proposed for the radical-scavenging by Bors
et al.11,12 are also important for singlet oxygen quenching in
flavonoids.
4.2. Correlation between Log kQ and Peak Oxidation
Potentials. In a previous work, the rate of quenching of 1O2 by
17 kinds of tocopherol derivatives, including R-, â-, γ-, and
δ-tocopherols, and five structurally related phenols has been
measured spectrophotometrically in ethanol at 35 °C.29 The
result indicates that the overall rate constants, kQ, increase as
the total electron-donating capacity of the alkyl substituents on
the aromatic ring increases. Log kQ was found to correlate with
their peak oxidation potentials, EP. Similar correlation was
observed for the biological hydroquinones and related com-
pounds.30
L
Figure 8. Plot of log kQ vs 1/λmax for flavonol, chrysin, apigenin,
rutin, quercetin, myricetin, naringenin, and taxifolin.
of OH groups substituted to the flavone skeleton. The plot of
log kQ vs inverse absorption maximum (1/λmaxL) is shown in
Figure 8. The good correlation (R ) 0.96) with a slope of -(2.33
( 0.30) × 10-4 cm-1 was observed for flavonoids except for
flavonol, indicating that flavonoids with larger λmaxL values show
1
faster O2 quenching rates.
4. Discussion
4.1. Structure-Activity Relationship of the 1O2 Quenching
Reaction by Flavonoids in Ethanol Solution. As shown in
Figure 1, the flavonoids except for naringenin and taxifolin are
OH derivatives of flavone, in which the aromatic A- and B-rings
are π-conjugated to each other by a 2,3-double bond. As listed
in Table 2, the kQ values obtained are less than <3 × 105 M-1
s-1 for flavone, 5.33 × 106 M-1 s-1 for flavonol, and 2.01 ×
107 M-1 s-1 for chrysin. The reactivity of flavone without an
OH group is very weak and almost negligible. Flavonol with a
3-OH group at the C-ring shows low reactivity. Chrysin has
higher reactivity than flavonol, clearly indicating that the
1
resorcinol A-ring contributes to the O2 quenching action by
Measurements of EP of flavonoids (chrysin, apigenin, rutin,
quercetin, naringenin, and taxifolin) were performed in 1:1 (v/
v) water-ethanol containing 50 mM KCl and 50 mM phosphate
buffer (pH 7.0) by Hotta et al.,31 using cyclic voltammetry. The
EP values reported are listed in Table 2. The values of log kQ
for flavonoids have been plotted against EP. As shown in Figure
9, log kQ of chrysin, apigenin, rutin, and quercetin (compounds
with 2,3-double bonds) correlates with EP with a slope of -2.2
flavonoids. The reaction rate (kQ ) 2.84 × 107 M-1 s-1) of
apigenin, with a 4′-OH group at the B-ring, is larger than that
of chrysin, suggesting that the 4′-OH group at the B-ring
contributes to the 1O2 quenching. Further, the kQ values obtained
for rutin (kQ ) 1.21 × 108 M-1 s-1) and quercetin (kQ ) 4.57
× 108 M-1 s-1), which have 3′- and 4′-OH groups at the B-ring,
are 6.0 and 23 times larger than that for chrysin. The result

Quenching of Singlet Oxygen by Flavonoids in Ethanol
J. Phys. Chem. B, Vol. 109, No. 9, 2005 4239
observed for the carotenoid derivatives having comparatively
higher excitation energies [(2.20∼2.30) × 104 cm-1], that is,
slower quenching rates [kQ ) (1.0 × 108)∼(3.0 × 109) M-1
s-1].35 The result suggests that flavonoids with smaller Eex values
show higher 1O2 quenching activities and, thus, higher biological
activity, as described above.
The results of the X-ray structure analyses show that the
torsion angles (θ) between B- and C-rings in flavone derivatives
without an OH substituent at the 3-position are 0.7∼24.1° (avg
12°), indicating that the molecules take planar structures.36
Similarly, the torsion angles (θ) in flavonol and quercetin with
an OH substituent at the 3-position are 5.5° and 7°, respectively,
indicating planar structure.37,38 Both quercetin and rutin, which
is quercetin rutinoside at the 3-position, have OH substituents
at 3′-, 4′-, 5-, and 7-positions, and we can expect similar rate
constants (kQ) for these flavonoids. However, the kQ value of
rutin is 3.8 times smaller than that of quercetin in ethanol, as
listed in Table 2. The π-conjugation between B- and C-rings in
rutin will be weaker than that in quercetin, because the B-ring
of rutin is considered to twist much more than that of quercetin
by the steric repulsion between the 6′- (or 2′-) ring proton at
the B-ring and the rutinose group. In such a case, the energy
level of HOMO (EHOMO) of rutin lowers, the oxidation potential
Figure 9. Plot of log kQ vs EP for chrysin, apigenin, rutin, quercetin,
naringenin, and taxifolin.
( 0.2 V-1 (correlation coefficient ) -0.99). The flavonoids
that have smaller EP values show higher reactivities. The result
1
suggests that the transition state in the above O2 quenching
reaction by flavonoids has the property of a charge-transfer
intermediate.32 On the other hand, the values of naringenin and
taxifolin, without 2,3-double bonds, deviate from the above
correlation line. The EP values reported for naringenin and
taxifolin are considered to be smaller than those expected from
L
(EP) of rutin increases, the λmax value decreases, and thus the
kQ value will decrease.29,30,35 In fact, the EP value of rutin (0.360
V vs Ag/AgCl) is larger than that of quercetin (0.178 V),31 and
the λmaxL value of rutin (361 nm) is smaller than that of quercetin
(371 nm).
EHOMO values obtained by MO calculation, as described below.
As reported in a previous work, the rate constants of
4.4. Comparison between the Quenching Rates (kQ) of
Flavonoids and Biological Compounds. Singlet oxygen reacts
with a wide variety of biological targets including lipids, sterols,
proteins (amino acids), DNA, etc. Rate constants kQ ()kq + kr)
for a large number of these reactions have been reported
scavenging of ArO• (kS) by tocopherol derivatives increase as
the total electron-donating capacity of the alkyl groups at the
aromatic ring increases.33 A plot of log kS and log kQ versus
peak oxidation potential (EP) was found to be linear and the
slope was negative. Linear correlation between the rates of
quenching of singlet oxygen (kQ) and scavenging of peroxyl
and aroxyl radicals (kS) in solution was found.29 As described
in a previous section, the kS and kR values of flavonoids increase
in the order shown in eq 3, independent of both pH value and
solvent system.17 On the other hand, the EP values of the above
flavonoids reported by Hotta et al. decrease in the order chrysin
> apigenin > flavonol > rutin > quercetin. Hendrickson et
al.34 found that the effect of flavonoids on microsomal phenol
hydroxylase activity correlates well with the oxidation potential
(EP) for flavonoids aglycons; the flavonoids that have smaller
EP values show higher inhibitions of phenol hydroxylase activity.
Further, the correlation between the EP values of flavonoids and
their log IC50 values for doxorubicin-induced lipid peroxidation
has been reported by Acker et al.6 These results suggest that
the flavonoids with smaller EP values show higher free radical
scavenging and singlet oxygen quenching activities, and thus
higher biological activity.
1
previously (see Table 1 in ref 39). Most reactions of O2 with
biological targets occur via chemical rather than physical routes.
For instance, peroxidation of unsaturated lipids is induced by
1
singlet oxygen. The quenching rates (kQ) of O2 by saturated
and unsaturated fatty acids and lipids are 9.0 × 103 M-1 s-1
for stearic acid, 1.7 × 104 M-1 s-1 for oleic acid, 4.2 × 104
M-1 s-1 for linoleic acid, and 6.0 × 104 M-1 s-1 for egg yolk
phosphatidylcholine.40,41 The quenching rate increases as the
number of double bonds in the fatty acid molecule increases.
The kQ values [(3.68 × 106)∼(5.12 × 108) M-1 s-1] observed
for eight kinds of flavonoids are 2-4 orders of magnitude larger
than those for fatty acids and phospholipid. The result suggests
1
that these flavonoids may contribute to the quenching of O2
and prevent lipid peroxidation in cell membranes. Similarly,
the kQ values observed for flavonoids are 1-3 orders of
magnitude larger than that (5.1 × 105 M-1 s-1) for DNA.
Davies et al.39,42 reported that proteins will be major targets
1
1
for O2 within cells, as the rate constants for reaction of O2
with amino acid side chains in proteins are higher than those
with most other cellular targets, and proteins are present at high
concentrations when compared to other species within cells. Of
the common amino acids present in proteins, only Try, His,
Tyr, Met, and Cys react at significant rates at physiological pH
values. Rate constants reported for these amino acids are
(0.8∼3.7) × 107 M-1 s-1. The values are similar to those of
chrysin (2.01 × 107 M-1 s-1) and apigenin (2.84 × 107 M-1
s-1) and 1-2 orders of magnitude smaller than those of rutin
(1.21 × 108 M-1 s-1), quercetin (4.57 × 108 M-1 s-1), and
myricetin (5.12 × 108 M-1 s-1).
4.3. Correlation of Log kQ with Energy Level of HOMO
and Inverse Absorption Maximum in UV-Vis Absorption
Spectra. As shown in Figure 6, the flavonoids that have higher
EHOMO values show higher reactivities. The result is reasonable,
because the flavonoids that have higher EHOMO values will show
smaller ionization potential (IP), that is, smaller oxidation
potential (EP).
UV-Vis absorption spectra of flavonoids with 2,3-double
L
bonds are shown in Figure 7. The λmax values for the longest
wavelength ππ* excitation in flavonoids increase with increasing
number of OH groups substituted to the flavone skeleton. As
shown in Figure 8, a good correlation between log kQ and
1/λmaxL, that is, the longest-wavelength ππ* excitation energy
(Eex), was observed for flavonoids. A similar correlation was
R-, â-, γ-, and δ-Tocopherol and biological hydroquinones,
such as ubiquinol-10 (UQ10H2), vitamin K1 hydroquinone, and

4240 J. Phys. Chem. B, Vol. 109, No. 9, 2005
Nagai et al.
(10) Rice-Evans, C. A.; Miller, N. J.; Paganga, G. Free Rad. Biol. Med.
1996, 20, 933-956.
(11) Bors, W.; Saran, M. Free Rad. Res. Commus. 1987, 2, 289-294.
(12) Bors, W.; Heller, W.; Michel, C.; Saran, M. Flavonoids as
antioxidants: Determination of radical-scavenging efficiencies. In Methods
in Enzymology; Academic Press: New York, 1990; Vol. 186, pp 343-
355.
(13) Jovanovic, S. V.; Steenken, S.; Tosic, M.; Marjanovic, B.; Simic,
M. G. J. Am. Chem. Soc. 1994, 116, 4846-4851.
(14) Jovanovic, S. V.; Hara, Y.; Steenken, S.; Simic M. G. J. Am. Chem.
Soc. 1995, 117, 9881-9888.
(15) Terao, J.; Piskula, M.; Yao, Q. Arch. Biochem. Biophys. 1994, 308,
278-284.
(16) Ioku, K.; Tsushida, T.; Takei, Y.; Nakatani, N.; Terao, J. Biochim.
Biophys. Acta 1995, 1234, 99-104.
(17) Mukai, K.; Oka, W.; Watanabe, K.; Egawa, Y.; Nagaoka, S. J.
Phys. Chem. A. 1997, 101, 3746-3753.
(18) Hideg, E.; Kalai, T.; Hideg, K.; Vass, I. Biochemistry 1998, 37,
11405-11411.
(19) Yasui, H.; Sakurai, H. Biochem. Biophys. Res. Commun. 2000, 269,
131-136.
plastoquinol (PQH2), are well-known as the most popular lipid-
soluble antioxidants. γ-Tocopherol hydroquinone (γ-TQH2) is
considered to be a plastoquinol model, because both PQH2 and
γ-TQH2 have two methyl substituents at 2- and 3-positions and
a long alkyl chain at the 6-position, and thus the rate constants
kQ of PQH2 and γ-TQH2 are thought to be similar to each other.
1
The quenching rates of O2 by these antioxidants have been
reported in previous works (see Table 2).29,30 As described in a
previous section, the rate of the quenching reaction of 1O2 with
flavonoids increases in the order shown in eq 6 in ethanol
solution. The rate constants (kQ) obtained for rutin (1.21 × 108
M-1 s-1), quercetin (4.57 × 108), and myricetin (5.12 × 108)
are similar to (or larger than) those of R-tocopherol (2.06 ×
108 M-1 s-1), γ-tocopherol (1.38 × 108), UQ10H2 (1.58 × 108),
and γ-TQH2 (plastoquinol model) (1.17 × 108). Flavonoids are
found in high concentration in foods and plants. The present
kinetic study suggests that the above flavonoids function as
singlet oxygen quenchers in biological systems (such as cell
membranes, photosynthetic systems, etc.) and protect the
systems from oxidative damage. However, the quenching rates
of these flavonoids are 1-2 orders of magnitude smaller than
that (1.58 × 1010 M-1 s-1) of â-carotene, which is well-known
(20) Tournaire, C.; Sylvie, S.; Maurette, M.-T. J. Photochem. Photobiol.
B: Biol. 1993, 19, 205-215.
(21) Saito, I.; Matsuura, T.; Inoue, K. J. Am. Chem. Soc. 1983, 105,
3200-3206.
(22) Inoue, K.; Matsuura, T.; Saito, I. Tetrahedron 1985, 41, 2177-
2181.
(23) Young, R. H.; Wehrly, K.; Martin, R. L. J. Am. Chem. Soc. 1971,
93, 5774-5779.
1
as a representative O2 quencher.
(24) Stewart, J. P. J. Comput. Chem. 1989, 10, 209-220.
(25) Merkel, P. B.; Kearns, D. R. J. Am. Chem. Soc. 1972, 94, 7244-
7253.
(26) Fahrenholtz, S. R.; Doleiden, F. H.; Trozzolo, A. M.; Lamola, A.
A. Photochem. Photobiol. 1974, 20, 505-509.
(27) Clough, R. L.; Yee, B. G.; Foote, C. S. J. Am. Chem. Soc. 1979,
101, 683-686.
(28) Gorman, A. A.; Gould, I. R.; Hamblett, I.; Standen, M. C. J. Am.
Chem. Soc. 1984, 106, 6956-6959.
(29) Mukai, K.; Daifuku, K.; Okabe, K.; Tanigaki, T.; Inoue, K. J. Org.
Chem. 1991, 56, 4188-4192.
(30) Mukai, K.; Itoh, S.; Daifuku, K.; Morimoto, H.; Inoue, K. Biochim.
Biophys. Acta 1993, 1183, 323-326.
(31) Hotta, H.; Nagano, S.; Ueda, M.; Tsujino, Y.; Koyama, J.; Osakai,
T. Biochim. Biophys. Acta 2002, 1572, 123-132.
(32) Thomas, M. J.; Foote, C. S. Photochem. Photobiol. 1978, 27, 683-
693.
Acknowledgment. We are very grateful to Professor Kenzo
Inoue of Ehime University for his kind help in the preparation
of endoperoxide (EP). We are grateful to Professor Shin-ichi
Nagaoka of Ehime University for his helpful discussions. We
are also grateful to Mr. Takayuki Ishikawa of Canon Inc. for
his kind help in the semiempirical PM3 MO calculation. This
work was partly supported by the Grant-in-Aid for Scientific
Research on Priority Areas “Applications of Molecular Spins”
(Area 769, Proposal 15087104) from the Ministry of Education,
Culture, Sports, Science and Technology (MEXT), Japan (to
K.M.).
References and Notes
(1) Hertog, M. G. L.; Feskens, E. G. M.; Lollman, P. C. H.; Katan, M.
B.; Kromhout, D. The Lancet 1993, 342, 1007-1011.
(33) Mukai, K.; Kageyama, Y.; Ishida, T.; Fukuda, K. J. Org. Chem.
1989, 54, 552-556.
(2) Hertog, M. G. L.; Kromhout, D.; Aravans, C.; Blackburn, H.;
Buzina, R.; Fidanza, F.; Giampaoli, S.; Jansen, A.; Menotti, A.; Nedeljkovic,
S.; Pekkarinen, M.; Simic, B. S.; Toshima, H.; Feskens, E. J. M.; Hollman,
P. C. H.; Katan, M. Arch. Intern. Med. 1995, 155, 381-386.
(3) Jankun, J.; Selman, S. H.; Swiercz, R.; Skrzypczak-Jankun, E.
Nature 1997, 387, 561.
(4) Mukhtar, H.; Katiyar, S. K.; Agarwal, R. Cancer chemoprevention
by green tea components. In Diet and cancer, markers, preVention, and
treatment; Jacobs, M. M., Ed.; Plenum Press: New York, 1994.
(5) Middleton, E., Jr.; Kandaswami, C. The impact of plant flavonoids
on mammalian biology: implications for immunity, inflammation and
cancer. In The flaVonoids: AdVances in research since 1986; Harbone, J.
B., Ed.; Chapman & Hall; London, 1993; 619-652.
(6) Acker, S. A. B. E.; Berg, D.-J.; Tromp, M. N. J. L.; Griffioen D.
H.; Bennekom, W. P.; Vijgh, W. J. F.; Bast, A. Free Rad. Biol. Med. 1996,
20, 331-342.
(34) Hendrickson, H. P.; Sahafayen, M.; Bell, M. A.; Kaufman, A. D.;
Hadwiger, M. E.; Lunte, C. E. J. Pharmaceut. Biomed. Anal. 1994, 12,
335-341.
(35) Beutner, S.; Bloedorn, B.; Frixel, S.; Blanco, I. H.; Hoffmann, T.;
Martin, H.-D.; Mayer, B.; Noack, P.; Ruck, C.; Schmidt, M.; Schulke, I.;
Sell, S.; Ernst, H.; Haremza, S.; Seybold, D.; Sies, H.; Stahl, W.; Walsh,
R. J. Sci. Food Agric. 2001, 81, 559-568.
(36) Kumar, S.; Ramanathan, T.; Subramanian, K.; Steiner, T. J. Chem.
Crystallogr. 1998, 28, 931-933.
(37) Etter, M. C.; Urbanczyk-Lipkowska, Z.; Baer, S.; Barbara, P. F. J.
Mol. Struct. 1986, 144, 155-167.
(38) Rossi, M.; Rickles, L. F.; Halpin, W. A. Bioorg. Chem. 1986, 14,
55-69.
(39) Davies, M. J.; Truscott, J. W. J. Photochem. Photobiol. B 2001,
63, 114-125.
(40) Krasnovsky, A. A.; Kagan, V. E.; Minin, A. A. FEBS Lett. 1983,
155, 233-236.
(41) Kohno, Y.; Egawa, Y.; Itoh, S.; Nagaoka, S.; Takahashi, M.; Mukai,
K. Biochim. Biophys. Acta 1995, 1256, 52-56.
(42) Davies, M. J. Biochem. Biophys. Res. Commun. 2003, 305, 761-
770.
(7) Santos, A. C.; Uyemura, S. A.; Lopes, J. L. C.; Bazon, J. N.;
Mingatto, F. E.; Curti, C. Free Rad. Biol. Med. 1998, 24, 1455-1461.
(8) Arora, A.; Nair, M. G.; Strasburg, G. M. Free Radical Biol. Med.
1998, 24, 1355-1363.
(9) De Whalley C. V.; Rankin S. M.; Hoult, R. S.; Jessup, W.; Leake,
D. S. Biochem. Pharmacol. 1990, 39, 1743-1750.