Kinetics of curcumin oxidation by 2,2-diphenyl-1-picrylhydrazyl (DPPH): An interesting case of separated coupled proton-electron transfer

Article abstract of DOI:10.1039/c6ob01439a

The decay of dpph in absolute ethanol at 25 °C and in the presence of curcumin (1), 4-methylcurcumin (3), 4,4-dimethylcurcumin (4) or curcumin 4′-methyl ether (5) follows bi-exponential kinetics. These unusual reaction kinetics are compatible with a two-step process in which an intermediate accumulates in a reversible first step followed by an irreversible process. As in other similar cases (Foti et al., Org. Lett., 2011, 13, 4826-4829), we have hypothesised that the intermediate is a π-stacked complex, formed between one curcumin anion (in the case of 1, 3 and 5 the enolate anion) and the picryl moiety of dpph, in which an intra-complex electron transfer from the (enolate) anion takes place. By comparing the kinetics of curcumin 4′,4′′-dimethyl ether (2) (no phenolic OH), (5) (one phenolic OH) and (1) (two phenolic OHs), we have deduced that the electron transfer process must be accompanied by a simultaneous proton transfer from the phenolic OHs to the bulk solvent (separated coupled proton-electron transfer). The rate constants k_α for the forward reaction of 2, 5 and 1 with dpph are in fact ～0, 7.5 × 10³ and 1.8 × 10⁴ M^-1 s^-1, respectively, in a clear dependence on the number of phenolic OHs.

Full text of DOI:10.1039/c6ob01439a

View Article Online
View Journal
Organic &
Biomolecular
Chemistry
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: M. C. Foti, A.
Slavova-Kazakova, C. Rocco and V. D. Kancheva, Org. Biomol. Chem., 2016, DOI: 10.1039/C6OB01439A.
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Accepted Manuscripts are published online shortly after
acceptance, before technical editing, formatting and proof reading.
Using this free service, authors can make their results available
to the community, in citable form, before we publish the edited
article. We will replace this Accepted Manuscript with the edited
and formatted Advance Article as soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes
to the text and/or graphics, which may alter content. The journal’s
standard Terms & Conditions and the Ethical guidelines still
apply. In no event shall the Royal Society of Chemistry be held
responsible for any errors or omissions in this Accepted Manuscript
or any consequences arising from the use of any information it
contains.
www.rsc.org/obc

Page 1 of 20
Organic & Biomolecular Chemistry
View Article Online
Kinetics of Curcumin Oxidation by 2,2-Diphenyl-1-Picrylhydrazyl
DOI: 10.1039/C6OB01439A
(DPPH•): an Interesting Case of Separated Coupled
Proton-Electron Transferα
Mario C. Foti,*,a Adriana Slavova-Kazakova,b Concetta Roccoa and Vessela D. Kanchevab
a Istituto di Chimica Biomolecolare del CNR, Via P. Gaifami 18 – I 95126 Catania (Italy);
b Institute of Organic Chemistry with Centre of Phytochemistry, Bulgarian Academy of Sciences,
“Acad. G. Bonchev” Str., Bl.9 - 1113 Sofia (Bulgaria)
α Electronic Supplementary Information (ESI) available: experimental details for the syntheses of
curcumin derivatives 2 – 5 together with their spectral characterization data (1H and 13C NMR
spectra).
Email: mario.foti@cnr.it
1

Organic & Biomolecular Chemistry
Page 2 of 20
View Article Online
DOI: 10.1039/C6OB01439A
Abstract
The decay of dpph• in absolute ethanol at 25 °C and in the presence of curcumin (1), 4-
methylcurcumin (3), 4,4-dimethylcurcumin (4) or curcumin 4’-methyl ether (5) follows bi-
exponential kinetics. These unusual reaction kinetics are compatible with a two-step process in
which an intermediate accumulates in a reversible first step followed by an irreversible process. As
in other similar cases (Foti et al. Org. Lett. 2011, 13, 4826-4829), we have hypothesised that the
intermediate is a -stacked complex, formed between one curcumin anion (in the case of 1, 3 and 5
the enolate anion) and the picryl moiety of dpph•, in which an intra-complex electron transfer from
the (enolate) anion takes place. By comparing the kinetics of curcumin 4’,4”-dimethyl ether (2) (no
phenolic OH), (5) (one phenolic OH) and (1) (two phenolic OHs), we have deduced that the
electron transfer process must be accompanied by a simultaneous proton transfer from the phenolic
OHs into the bulk solvent (separated coupled proton-electron transfer). The rate constants kfor the
forward reaction of 2, 5 and 1 with dpph• are in fact ~ 0, 7.5×103 and 1.8×104 M-1s-1, respectively,
in a clear dependence on the number of phenolic OHs.
2

Page 3 of 20
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C6OB01439A
Introduction
Curcumin (1,7-bis(4-hydroxy-3-methoxyphenyl)-1,6-heptadiene-3,5-dione) is a natural
phenolic compound (see Scheme 1) extracted from the rhizome of Curcuma longa (turmeric).1 India
is the major world producer of turmeric – especially in the regions of Madras and Alleppey, from
which the turmeric found on the market takes the name.1 The content of curcumin in the two
products is significantly different since the Alleppey turmeric contains about 4 – 7 % of curcumin
while the Madras type only 2%.1 However, several syntheses make now curcumin readily
available.2 The toxicity of curcumin seems to be very limited in humans because quantities as large
as 8 – 12 g/day were used in phase I clinical trials without adverse effects.3 The bioavailability of
curcumin is however poor since 8 g/day produce a peak serum concentration of only 1.77 M after
about 2 hours from oral intake.3,4
The scientific community has claimed5,6 that curcumin has a wide spectrum of health
benefits including anticancer, antioxidant, anti-inflammatory, anti-bacterial, wound healing and
hepatoprotective activities, though divergent opinions7 have been expressed. The poor absorption
and water solubility, however, seem to be a major obstacle to the therapeutic use of curcumin by
oral administration.8 This has encouraged the search for practical solutions in order to ameliorate
the oral absorption. In this respect, several strategies – mostly concerning nanoparticle-based drug
delivery – are now being investigated.8-10
Curcumin both in solution and solid state11,12 can have two proton-shift tautomeric forms:
the planar keto-enol form 1a and the nonplanar -diketo form 1b/1b’ represented in Scheme 1.
NMR experiments11 as well as quantum chemistry calculations13 show that the equilibrium is,
however, essentially entirely shifted toward the keto-enol tautomer. This is because of the
additional intramolecular stabilization through hydrogen bonding and electron delocalization in the
conjugated four-ene system present in 1a.13 The position of the tautomeric equilibrium is poorly
3

Organic & Biomolecular Chemistry
Page 4 of 20
11
View Article Online
DOI: 10.1039/C6OB01439A
influenced by the solvent. However, in DMSO-d6 at 371 K substantial amounts (~10%) of the -
diketo curcumin, 1b, were observed by NMR (data not published).14,15
Scheme 1. Structures of the two proton-shift tautomers of curcumin: the keto-enol 1a (the other
identical keto-enol tautomer with shifted H-bond and double bonds is not represented) and the -
diketone 1b (1b’ is probably less important because of the repulsion between the two carbonyls, see
ref. 13). The keto-enol form is, however, essentially the sole tautomer that the NMR spectroscopy
detects at room temperature in solutions of curcumin ~ 10-3 M and in solvents ranging in polarity
from CDCl3 to DMSO-d6/D2O mixtures, see ref. 11. However, in DMSO-d6 at 371 K, the -diketo
form was observed and the equilibrium constant Keq = [1b]/[1a] was evaluated to be ca. 0.12.
Scheme 2. Curcumin derivatives 2 – 5 studied in the present work.
4

Page 5 of 20
Organic & Biomolecular Chemistry
View Article Online
To investigate the influence of the keto-enol moiety over the mechanismDOoIf: 10r.e10a3c9t/iCo6nOB0o1f439A
curcumin with dpph• we synthesized two derivatives with one or two methyl groups at the central
position 4 (compounds 3 and 4 in Scheme 2). Compound 4 is obviously exclusively present as a -
1
diketone. In the case of compound 3, we observed from the H NMR spectrum that the keto-enol
tautomer 3a was in equilibrium with large quantities of the -diketo form 3b in solution of CDCl3 at
room temperature (molar fraction 푥ퟑ풃 ≅ 0.29), see the Electronic Supplementary Information
13
(ESI). The C NMR spectrum confirmed the presence of the two tautomers in equilibrium for the
presence of signals at 196.06 (C=O of -diketo 3b), 182.39 (C=O of keto-enol 3a), 13.07 (CH3 at
C-4 in 3b) and 11.99 ppm (CH3 at C-4 in 3a), see ESI.
The reason as to why 3b is present in such large quantities might be due to hyperconjugation
or steric hindrance between the methyl group and the two hydrogen atoms of the adjacent double
bonds, which might restrict the internal rotation of the CH3 group in the planar keto-enol tautomer
3a.16 On the contrary, in 3b the methyl internal rotation is essentially free:
The site of reaction of curcumin with free radicals has been the subject of a (scientifically
exciting) controversy.17,18 Jovanovic and co-workers, in one initial work on the reactions of
curcumin with free radicals, claimed that the methyl and tert-butoxyl radicals reacted in acidic
aqueous solutions (the methyl) and acetonitrile (the tert-butoxyl) at the central CH2 group of the
heptadienone link (they hypothesized that curcumin was essentially present as the diketo form).18
5

Organic & Biomolecular Chemistry
Page 6 of 20
18
View Article Online
However, at pH > 8 the phenolic part of curcumin would take over as electron-donor reaction site.
DOI: 10.1039/C6OB01439A
According to Barclay and co-workers,17 peroxyl radicals reacted instead with curcumin in
chlorobenzene by H-atom abstraction from the phenolic groups not the CH2 group. This
discrepancy was actually resolved in 2004 when Litwinienko and Ingold19a proposed a different
interpretation of the data by simply invoking solvent effect and sequential proton loss electron
transfer (SPLET) mechanism.19
The antioxidant ability of biophenols is frequently (and erroneously20) evaluated from their
ability to reduce the dpph• radical in protic solvents like methanol or ethanol.21 This reducing
ability, however, differs significantly from the chain-breaking activity of antioxidants, that is, from
their ability to shorten the oxidation chain-length of oxidizable substrates.22 The electron-deficient
radical dpph• reacts in ionizing solvents (alcohols and their mixtures with water) with phenols
essentially via electron transfer (ET) processes and with mechanisms that may contemplate the
formation of a -stacked complex.23 With relatively smaller electron deficiency, ordinary peroxyl
radicals react instead via H-atom abstraction22 and thus, the comparison between these two
reduction mechanisms can be misleading.20
In this work, we intend to analyse the kinetics of the curcumin/dpph• reaction in ethanol, in
order to “decrypt” the mechanism involved, by using the insights given by Litwinienko and Ingold
in their noticeable work.19a In ethanol, as well as in methanol, this reaction did not follow ordinary
pseudo-first-order kinetics despite in most of our experiments [curcumin] ≫ [dpph•]. This implies
that this reaction is not an elementary bimolecular reaction, a + b  c. We found that these
remarkable but not unique kinetics were in fact well-fitted by a bi-exponential function, and this
supported the idea that the reaction occurred in a two-step process in which an intermediate (-
stacked complex, vide infra) accumulated in a reversible step.24-27
On these grounds, we propose that the reaction of dpph• with curcumin and some curcumin
derivatives (see Scheme 2) occurs in ethanol via ET from the curcumin enolate to dpph• with the
6

Page 7 of 20
Organic & Biomolecular Chemistry
View Article Online
reaction mechanism represented below. By analysing the reactivity of 1, 2 and 5 we made a most
DOI: 10.1039/C6OB01439A
intriguing observation: the ET process is accompanied by a simultaneous release of a proton from
the phenolic OHs into the bulk solution.
Experimental
Technical grade curcumin was purchased from Aldrich and was purified before use by
chromatography over LiChroprep 40 – 63 m with a gradient of acetone (1% to 5%) in
dichloromethane. The dpph• radical was purchased from Sigma-Aldrich and was used without
further purification. Absolute ethanol for HPLC analysis was purchased from Fluka. Lipozyme® 42
U/g, immobilized from Mucor miehei and used to hydrolyse the acetylated form of compound 3,
was obtained from Fluka. The syntheses of compounds 2 – 5 and their spectral characterization data
are reported in ESI.
Kinetics with dpph•. The kinetics was measured with a Hi-Tech Scientific stopped-flow
spectrophotometer connected to a Julabo thermal bath set at 25 °C. Stock solutions of purified
curcumin or the other compounds 2 – 5 ca. (1 – 2) mM were prepared in absolute ethanol and kept
in a refrigerator until use. The dpph• solution (≤ 1 mg/10 mL) was freshly prepared in ethanol with
7

Organic & Biomolecular Chemistry
Page 8 of 20
View Article Online
an initial absorbance ≤ 2 at 517 nm, in such a way that in the stopped-flow speDcOtrI:o1p0.h10o39to/Cm6OeBt0e1r439A
(solution diluted 1:1) the absorbance was ≤ 1. The solution of antioxidants was properly diluted to
concentrations of the order of 10 – 1000 M. Once the syringes of the instrument were filled with
the antioxidant and dpph• solutions and thermal equilibrium was achieved, the solutions were
mixed and the absorbance decay at 517 nm was recorded. Each single rate constant at a given
concentration of antioxidant was obtained by averaging the values obtained from three to six decay
traces recorded within a time scale of 1 – 2 seconds. The observed deviations from the average were
less than ± 20%. The scheme of reaction represented in reaction 2 along with the concentrations of
antioxidant and dpph• were introduced in the program SPECFIT which, by simulation of the
experimental decay traces absorbance vs. time, afforded the rate constants involved in the reaction
scheme. These rate constants were essentially independent of the concentrations of curcumin and
curcumin analogues 2 – 5.
Results and Discussion
In our experiments (as well as in others’)19a, the decay of dpph• (≤ 80 M) in the presence
of curcumin (20 – 2000 M) (or curcumin derivatives 3 – 5) occurred in ethanol with biphasic
kinetics, see Figure 1. Curve fitting of the kinetics with [curcumin] >> [dpph•] by first-order mono-
exponential functions (i.e., At = A0 × e-kt) was usually very poor. Excellent results (see Figure 1)
were instead obtained by applying a bi-exponential function as equation 3,
퐴푡 = 푎 × 푒−훿 푡 + 푏 × 푒−훿 푡 + 푐
(3)
1
2
where At is the dpph• absorbance (at 517 nm) monitored over time t and a + b + c = A0. Two pseudo
rate constants, 1 and 2, were therefore obtained for each concentration of curcumin by fitting the
experimental curves with equation 3. In all cases, one pseudo rate constant, say, 1, was larger than
8

Page 9 of 20
Organic & Biomolecular Chemistry
View Article Online
the other one 2 by ca. one-order of magnitude, i.e. 1 > 2. Furthermore, the two pseudo-rate
DOI: 10.1039/C6OB01439A
constants increased as the curcumin concentration was increased while no dependence over the
dpph• concentration was observed.
0.94
1.25
0.92
1.05
0.90
0.88
0.86
0.84
0.82
0.80
0.78
0.85
0.65
0.45
0.25
0
0.2
0.4
0.6
TIME / s
0.8
1
1.2
0
2
4
6
8
10
12
Time / s
Figure 1. Examples of dpph• bleaching in the presence of curcumin in ethanol. The solid blue
circles represent the experimental absorbance readings for 1 + dpph• while the red line is the best
fitting line obtained with a bi-exponential function (equation 3). The concentrations of curcumin
and dpph• were 1.54 × 10-5 M and 8.4×10-5 M, respectively. Inset: experimental readings (blue
circles) and bi-exponential fitting curve (in red) for [curcumin] = 8.38×10-4 M and [dpph•] = 8.4 ×
10-5 M in ethanol at 25 °C.
Equation 3 is compatible with several (similar) mechanisms24 but the most likely one is the
mechanism represented above in reaction 2. This mechanism contemplates the formation of a -
stacked complex between dpph• and the enolate 1- accompanied by intra-complex ET as previously
formulated for other similar cases.23a Nonetheless, snapshots of the solution did not show any
9

Organic & Biomolecular Chemistry
Page 10 of 20
•
View Article Online
DOI: 10.1039/C6OB01439A
absorbing species other than dpph possibly because the stationary concentration of complex and
dpph anion were too small, see reaction 2. At any rate, we have noticed that a similar mechanism
had recently been formulated for the reaction of electron-rich phenols with peroxyl radicals in
aqueous solutions and this supports our mechanistic framework.28
Two different strategies were used to determine the rate constants k1, k-1 and k2 of reaction 2.
We first tackled the problem by using the program SPECFIT which tailored the rate constants in
order to fit the experimental kinetics. Because we had introduced in the program the concentration
̶
of curcumin and not that (unknown) of its anion 1 , the simulation did not give the rate constant k1
푲
푎
(see reaction 2) but the value 푘α = 푘1 ×
where Ka is the acidity constant of the keto-enol
+
[H ]
푠표푙
moiety (the most acidic group in curcumin; in methanol:water 1:1 v/v, pKa = 8.54)19a and [H+]sol is
the acidity of the ethanol in use, which was relatively large and uninfluenced by the addition of
small amounts of curcumin.29 In Table 1, the rate constants 푘α, k-1 and k2 obtained by SPECFIT are
reported.
The other strategy was based on the mathematical analysis of the rate-laws for a simplified
version of the mechanism represented in reaction 2. In first-order conditions (i.e. [curcumin] ≫
푲
푎
[dpph•]) and constant solvent acidity, the [1-] =
×[CurcOH] does not change appreciably
+
[H ]
푠표푙
over the course of the reaction. The forward reaction 2 can therefore be treated as a unimolecular
퐾
푎
′
−
[
]
reaction with a pseudo-first order rate constant 푘1 = 푘1 × CurcO = 푘1 × [H
[CurcOH]), and
+
]
푠표푙
the stepwise reaction can be represented as in reaction 4 where I is the intermediate -complex,
The integration of the differential equations for reaction 4 is complex (it requires Laplace
transforms or other mathematical techniques).24-27 The exact solutions show that the concentration
10

Page 11 of 20
Organic & Biomolecular Chemistry
View Article Online
of the three species involved varies in a bi-exponential manner over time as experimentally
DOI: 10.1039/C6OB01439A
observed for dpph• (see above and Figure 1). They also show that the pseudo-rate constants 1 and
2 of eq. 3 are related to the rate constants of reaction 4 through the relationships 5 and 6,27
1 + 2 = k’1 + k-1 + k2
1×2 = k’1 × k2
(5)
(6)
Equation 5 can be rewritten as,
1 + 2 = k [CurcOH] + k0
(7)
퐾
푎
where k = 푘1 × [H
(see above) and k0 = k-1 + k2. Hence, a plot of 1 + 2 against [CurcOH] must
+
]
푠표푙
afford a straight-line with slope equal to k and intercept k0. Figure 2 shows that the reaction of
curcumin (and curcumin derivatives 3 – 5) with dpph• satisfied this expectation. From these plots,
the value of k/M-1s-1 was determined to be ~ 1.8×104 for 1; 3.4×104 for 3; 7.5×103 for 5; 2×103 for
4. The mean value of the rate constant k2 was obtained from eq. 6 using the values of k’1 at different
[CurcOH]. Finally, k-1 was calculated from the intercept k0 = k-1 + k2. The rate constants so obtained
are in good agreement with those obtained via kinetic simulation with SPECFIT, see Table 1.
11

Organic & Biomolecular Chemistry
Page 12 of 20
View Article Online
40
30
20
10
0
DOI: 10.1039/C6OB01439A
0.0
0.3
0.6
0.9
1.2
1.5
1.8
[curcumin] / mM
Figure 2. Plot of the pseudo rate constants of curcumin, 1 + 2, at 25 °C vs. [curcumin]. The
linearity of the plot is good (R2 = 0.98) and the slope affords the value of k×104 M-1s-1. The
intercept gives k0 = 6.4 s-1. The error was about ± 20%.
Table 1. Rate constants relative to the reaction of dpph• with 1 – 5 in absolute ethanol at 298 K
(error ± 20%). The values were determined with two different procedures, see text; the data
obtained with the program SPECFIT are reported in italics while those obtained by eqs. 5 – 7 are in
plain characters.
Compound
kM-1s-1)
ks-1)
ks-1)
(1.8 ± 0.5)×104
3 ± 1
3.4 ± 1.0
(2.4 ± 1.5)×104
1.1 ± 0.9
0.4 ± 0.2
1
2
3
4
0.004a
0
0
(3.4 ± 0.9)×104
(3.0 ± 1.5)×104
8 ± 2
2.8 ± 0.8
6.0 ± 1
2.0 ± 0.7
(1.8 ± 0.9)×103
2 ± 1
3 ± 1
12

Page 13 of 20
Organic & Biomolecular Chemistry
(2.0 ± 1.5)×103
2.0 ± 0.5
0.2b
View Article Online
DOI: 10.1039/C6OB01439A
(7.5 ± 1.5)×103
3.3 ± 1.5
4.7 ± 1.5
(5.5 ± 1.5)×103
5
2.0 ± 0.5
3.1 ± 0.5
aFirst-order rate constant (units in s-1) essentially independent of the concentration of 2 in the range 20 – 470
b
M. In these simulations the value of k2 was fixed at 0.2 s-1 to allow the convergence of the statistical
procedure.
Curcumin has two reaction sites: the enolic OH and the two phenolic OHs which are
selectively activated by solvent effects.30 Barclay and coworkers17 unequivocally demonstrated that
in styrene/chlorobenzene the phenolic OHs were the target for ROO• radicals which reacted by H-
atom abstraction with a rate constant of 3.4×105 M-1s-1 at 30 °C. This was deduced from the fact
that non-phenolic curcuminoids (e.g. 2 in Scheme 2) were unable to react with ROO• radicals and
therefore did not possess any antioxidant activity.17 On the other hand, in ionizing solvents (CH3OH
and C2H5OH), Litwinienko and Ingold demonstrated that the reagent involved in the reaction with
dpph• was one curcumin anion and this was demonstrated through the effect of added acetic acid.19a
The reaction rate declined in a dose-dependent manner since the ionization of curcumin, and hence
the anion, at sufficiently high [CH3COOH] was suppressed. That this anion was the enolate was
deduced from the pKas of curcumin relative to compound 2. In methanol/water 1:1 v/v, the pKa, of
2, due to the keto-enol moiety, is 8.75.19a The pKas of curcumin are 8.54, 9.30 and 10.69 and thus, it
is evident by comparison that the lowest pKa is due to the enolic OH which is therefore the most
acidic group in curcumin.19a
The authors also gave a plausible justification of the origin of the biphasic kinetics. They
attributed the fast initial decay to the reaction of dpph• with preformed curcumin anions present at
low equilibrium concentration. Once these initial anions were depleted, the slow ionization of
13

Organic & Biomolecular Chemistry
Page 14 of 20
curcumin became partly rate-determining generating the “stretched” exponential observedVie.1w9AarticTleoOnline
DOI: 10.1039/C6OB01439A
confirm or reject this reasonable hypothesis we synthesized two curcumin analogues, 2 and 5,
which were present in solution essentially entirely in the keto-enol form as deduced from their
proton and carbon NMR spectra, see ESI. The quantity of preformed anions from the ionization of 2
or 5 must be very similar to that of curcumin since the pKas of the keto-enol moieties of these
species are likely to be very similar (e.g. 8.75 for 2 vs. 8.54 for 1, see above). With these premises,
the reactivity of these three species was expected to be similar. In truth, we discovered that these
species behaved very differently. The curcumin analogue 2 reacted with a first-order rate constant
of only 0.004 s-1 that was essentially independent of the concentration (range 20 – 470 M). On the
other hand, the analogue 5 reacted with a kof ×M-1s-1, essentially equal to half the rate
constant of curcumin 1 (see Table 1). This different reactivity rules out the explanation given above
for the kinetic behaviour of curcumin attributed to the depletion of the enolate anions. We therefore
think that the biphasic kinetics is due to a two-step reaction as exemplified in reaction 2.
The reactivity of 1, 2 and 5 together with the ratio 푘훼(ퟏ)/푘훼(ퟓ) ≈ 2 suggest another
important aspect of the electron transfer reaction between enolate and dpph•. The ET process in 1
and 5 appears to be synchronized to a proton release from the phenolic OH into the bulk solution
̶
̶
with the consequent formation of the radical anions 1• or 5• , see reaction 2.31 The measured rate
constants 푘훼 are in fact dependent on the number of phenolic OHs present in the molecule, see
Table 2. This correlation suggests that the PT process cannot occur after the reduction of dpph•
because in this case it would be kinetically irrelevant.32 The alternative is that ET and PT occur in
concert (one single step), that is, the reaction between the enolate of 1 and 5 and dpph• is a special
case of separated coupled proton-electron transfer (separated CPET).33
14

Page 15 of 20
Organic & Biomolecular Chemistry
View Article Online
Table 2. Correlation between number of phenolic OHs and rate constant for the (intDrOacI: o10m.10p3l9e/Cx6)OEB0T1439A
from enolate anion to dpph• for curcuminoids present in solution exclusively in the keto-enol form.
Experimental error ca. ± 20%.
Compound
Number of Phenolic OHs
Rate Constant k
18.0 × 103 M-1s-1
7.5 × 103 M-1s-1
~ 0
2
1
0
1
5
2
The reactivity of the curcumin analogues 3 and 4 shows the importance that the keto-enol
moiety has in the reaction of curcuminoids with dpph•. Compound 4 has evidently no enol form
while for 3 the proton NMR spectrum, see ESI, shows that both tautomeric forms are present in
[
]
solution (the equilibrium constant was determined to be 푲eq = ퟑ풃 /[ퟑ풂] ≅ 0.4 in CDCl3 at room
temperature).
Ionization of compound 4 occurs at the phenolic moieties and the phenolate anion produced
can react by ET to dpph•, reaction 2. The rate constant is significantly lower than that of curcumin
1 (see Table 1) because 4 is probably less acidic than curcumin. In the case of 3, (which is present
in CDCl3 solution for 71% as the enol form 3a), the k was 33600 M-1s-1 (see Table 1). Given that
the di-keto form is likely to be much less reactive than the enol (e.g., kfor 4 is only 1800 M-1s-1,
see Table 1), the k for 3a, in the hypothesis that even in EtOH 3a is present for 71%, can be
estimated to be k(3a) = 33600/0.71 = 47300 M-1s-1. This means that the enol form of 3 would
roughly be 24-fold more reactive than 4 and ca. 3-fold than curcumin. The higher reactivity of 3
with respect to curcumin might be due to the fact that the methyl group can lower the ionization
potential and thus the ET free energy of the process. Equivalently, the methyl group can have a
stabilizing effect on the incipient enolic radical 3a•,
15

Organic & Biomolecular Chemistry
Page 16 of 20
View Article Online
DOI: 10.1039/C6OB01439A
Conclusions
Curcumin 1 and curcumin derivatives 3 and 5 react with dpph• in ethanol by electron-
transfer from the enolate anions in a -complex formed by the latter with the picryl portion of
dpph•. The observed bi-exponential kinetics implies a stepwise reaction with a reversible step (-
complex formation) followed by an irreversible step (-complex dissociation to the final products).
The analysis of the kinetic data supports the view that the ET process must be accompanied by a
simultaneous proton release from the phenolic OHs into the bulk solution (separated CPET) with
̶
̶
̶
the formation of relatively stable radical anions, 1• , 3• and 5•. 4-Methylcurcumin 3, which is present
in solution as a mixture of the two possible tautomers, is the most efficient curcuminoid compound
in the bleaching of dpph• with a rate constant k for the forward reaction of 33600 M-1s-1 (47300 M-
1s-1 for 100% enol tautomer). The above mechanism may have a wider application for it has
recently been applied to the reaction of electron-rich phenol with peroxyl radicals in aqueous
solutions, see ref. 28.
Acknowledgements
We are grateful to all four anonymous reviewers of this work for their constructive and pertinent
observations. Furthermore, we are indebted to Prof. Gregory Litwinienko (University of Warsaw),
Dr. Keith U. Ingold (National Research Council of Ottawa) and Prof. Søren T. Barsberg (University
of Copenhagen) for having read the manuscript and for their relevant comments.
16

Page 17 of 20
Organic & Biomolecular Chemistry
View Article Online
DOI: 10.1039/C6OB01439A
References and Notes
1) TURMERIC Post-harvest Operations (FAO) at http://www.fao.org/3/a-ax446e.pdf
2) a) H. J. J. Pabon Recl. Trav. Chim. Pays-Bas, 1964, 83, 379 – 386; b) C. E. Nichols,
D.Youssef, R. G. Harris, A. Jha, Arkivoc, 2006, 13, 64-72; c) E. V. Rao and P. Sudheer
Indian J. Pharm. Sci., 2011, 73, 262–270; d) K. I. Priyadarsini Molecules, 2014, 19,
20091-20112.
3) A. L. Cheng, C. H. Hsu, J. K. Lin, M. M. Hsu, Y. F. Ho, T. S. Shen, J. Y. Ko, J. T. Lin,
B. R. Lin, W. Ming-Shiang, H. S. Yu, S. H. Jee, G. S. Chen, T. M. Chen, C.A. Chen, M.
K. Lai, Y. S. Pu, M. H. Pan, Y.J. Wang, C. C. Tsai, C.Y. Hsieh Anticancer Res., 2001,
21(4B), 2895-2900.
4) P. Anand, A. B. Kunnumakkara, R. A. Newman, B. B. Aggarwal Mol. Pharmaceutics,
2007, 4, 807-818.
5) The Molecular Targets and Therapeutic Uses of Curcumin in Health and Disease; B. B.
Aggarwal, Y-J. Surh, S. Shishodia Eds.; Advances in Experimental Medicine and
Biology vol. 595; Springer: 2007.
6) B. B. Aggarwal and K. B. Harikumar, Int. J. Biochem. Cell Biol. 2009, 41, 40–59.
7) E. Burgos-Moron, J. M. Calderon-Montano, J. Salvador, A. Robles and M. Lopez-
Lazaro, Int. J. Cancer, 2010, 126, 1771–1775.
8) a) R. K. Basniwal, H. S. Buttar, V. K. Jain, N. Jain, J. Agric. Food. Chem. 2011, 59,
2056-2061; b) R. A. Sharma, W. P. Steward, A. J. Gescher, Adv. Exp. Med. Biol., 2007,
595, 453 – 470.
9) N. S. Rejinold,; R. G. Thomas, M. Muthiah, K. P. Chennazhi, K. Manzoor, I-K. Park, Y.
Y. Jeong, R. Jayakumar, Int. J. Biol. Macromol. 2015, 74, 249-262 and references
therein.
17

Organic & Biomolecular Chemistry
Page 18 of 20
View Article Online
10) E. C. Goethals, A. Elbaz, A. L. Lopata, S. K. Bhargava, V. Bansal, LangDmOuI:i1r0,.1023091/C36,OB20914,39A
658–666.
11) F. Payton, P. Sandusky, W. L. Alworth, J. Nat. Prod. 2007, 70, 143-146.
12) a) H. H Tønnesen, J. Karlsen, A.Mostad Acta Chem. Scand. Ser. B 1982, 36, 475-479; b)
J. T. Mague, W. L. Alworth, F. L. Payton, Acta Crystallogr. 2004, C60, 608-610.
13) K. Balasubramanian J. Agric. Food Chem. 2006, 54, 3512-3520.
1
14) The H NMR spectrum of a solution of curcumin in DMSO-d6 showed a weak but
visible signal at 4.1 ppm attributable to the CH2 of 1b (~ 3%) which increased as the
temperature was gradually raised to 371 K. On the other hand, the 13C spectrum showed
a signal at  195.6 attributable to the C=O of 1b. Several heteronuclear correlations
referable to 1b were also observed in the HMBC spectrum.
15) Addition of surfactants to aqueous solutions of curcumin stabilizes the -diketone form,
see: (a) A. Dutta, B. Boruah, P. M. Saikia, R. K. Dutta, J. Mol. Liq. 2013, 187, 350–358.
Large molar fractions of -diketo curcumin were found in ethanol/water at room
temperature by UV-vis spectroscopy at curcumin concentrations of ~ 10-5 M, see: (b) Y.
Manolova, V. Deneva, L. Antonov, E. Drakalska, D. Momekova, N. Lambov,
Spectrochim. Acta A 2014, 132, 815-820. This seems in disagreement with the NMR
data reported in ref. 11 and with our own NMR data. However, we observe that the
concentration used in the NMR experiments (~ 10-2 M) is much larger than that used in
the UV-vis spectroscopy (~ 10-5 M) and the equilibrium position may depend on the
curcumin concentration. The keto/enol equilibrium of acetylacetone shows in fact this
dependence in several solvents, see: (c) C. Reichardt in Solvents and Solvent Effects in
Organic Chemistry, third edition; WILEY-VCH Verlag GmbH & Co: Weinheim, 2003;
pp. 106-108.
18

Page 19 of 20
Organic & Biomolecular Chemistry
View Article Online
16) a) M. Baba, K. Mori, M. Saito,Y. Kowaka,Y. Noma, S. Kasahara, T. DYOIa: 1m0.1a0n39a/kCa6O, BK01.439A
Okuyama, T. Ishimoto, U. Nagashima, J. Phys. Chem. A 2009, 113, 2366–2371; b) H.-G.
Korth and P. Mulder, J. Org. Chem., 2015, 80, 8206–8216; c) K. T. Smith, S. C. Young,
J. W. DeBlasio, and C. S. Hamann, J. Chem. Educ. 2016, 93, 790 – 794.
17) L. R. C. Barclay, M. R. Vinqvist, K. Mukai, H. Goto, Y. Hashimoto, A. Tokunaga, H.
Uno, Org. Lett. 2000, 2, 2841-2843.
18) S. V. Jovanovic, S. Steenken, C. W. Boone, G. Simic, J. Am. Chem. Soc. 1999, 121,
9677-9681.
19) a) G. Litwinienko, K. U. Ingold, J. Org. Chem. 2004, 69, 5888-5896 and refs. therein; b)
M. C. Foti, C. Daquino, C. Geraci, J. Org. Chem. 2004, 69, 2309–2314; c) G.
Litwinienko and K. U. Ingold J. Org. Chem. 2003, 68, 3433-3438; d) G. Litwinienko
and K. U. Ingold, Acc. Chem. Res. 2007, 40, 222 – 230.
20) M. C. Foti, J. Agric. Food Chem. 2015, 63, 8765 – 8776.
21) W. Brand-Williams, M. E. Cuvelier, C. Berset, LWT Food Sci. Technol. 1995, 28, 25-30.
22) K. U. Ingold, D. A. Pratt, Chem. Rev. 2014, 114, 9022-9046.
23) a) M. C. Foti, C. Daquino, G. A. DiLabio, K. U. Ingold, Org. Lett. 2011, 13, 4826-4829;
b) R. Amorati, S. Menichetti, C. Viglianisi, M. C. Foti, Chem. Commun., 2012, 48,
11904–11906; c) M. C. Foti, C. Daquino, Chem. Commun. 2006, 3252–3254; d) Foti M.
C. Int. J. Chem. Kinet. 2012, 44, 524–531.
24) C. Aymard, A. Belarbi, Enzyme Microb. Technol. 2000, 27, 612–618.
25) A. R. Fersht and W. P. Jencks, J. Am. Chem. Soc. 1970, 92, 5432-5442.
26) B. H. Havsteen, J. Biol. Chem. 1966, 242, 769-776.
27) J. H. Espenson in Chemical Kinetics and Reaction Mechanisms, second edition,
McGRAW-HILL INTERNATIONAL EDITIONS 1995.
28) R. Amorati, A. Baschieri, G. Morroni, R. Gambino, and L.Valgimigli Chem. Eur. J.
2016, 22, 7924 – 7934.
19

Organic & Biomolecular Chemistry
Page 20 of 20
View Article Online
29) The titration of 50 mL of ethanol with aqueous 0.1 M NaOH in the presence of
DOI: 10.1039/C6OB01439A
phenolphthalein required 240 L of the base for colour change.
30) We have omitted the methylene group CH2 of 1b from our analysis for two reasons: i)
low equilibrium concentration of 1b ~ 3%; and ii) H-atom abstraction by ROO• or dpph•
from C atoms is usually slow, see ref. 22.
31) The most stable form of the radical anions 1•- and 5•- is that depicted in reaction 2 with
the unpaired electron in the phenolic moiety and the negative charge in the keto-enol
moiety, see ref. 19a and for the explanation see, M. C. Foti, R. Amorati, G.F. Pedulli, C.
Daquino, D. A. Pratt, K. U. Ingold J. Org. Chem. 2010, 75, 4434–4440.
32) It is highly improbable that the proton can be released before the intra-complex ET
because this would generate di-anions whose concentration in absolute ethanol is very
small.
33) a) J. M. Mayer, D. A. Hrovat, J. L. Thomas, W. T. Borden, J. Am. Chem. Soc. 2002,
124, 11142 – 11147; b) J. J. Warren and J. M. Mayer Biochemistry 2015, 54,
1863−1878.
20