Kinetics of ultra-fast excited state proton transfer from 7-hydroxy-4-methylflavylium chloride to water

Article abstract of DOI:10.1016/S0009-2614(98)01199-3

Excited state proton transfer from 7-hydroxy-4-methylflavylium chloride to water is reported. From a modified analysis of picosecond time-resolved fluorescence data (not using the lifetime of a parent compound), all rate constants were determined: the deprotonation rate constant of the flavylium cation, k_d=1.4×10¹¹ s^-1, the protonation rate constant of the base form, k_p=2.3×10¹⁰ l mol^-1 s^-1 and the reciprocal fluorescence lifetimes of these species, k_AH⁺=7.8×10⁹ s^-1 (τ_AH⁺=128 ps) and k_A=7.6×10⁹ s^-1 (τ_A=132 ps), in water, at 20°C. The value of k_d is the largest measured value for an intermolecular proton transfer (to water).

Full text of DOI:10.1016/S0009-2614(98)01199-3

11 December 1998
Ž
.
Chemical Physics Letters 298 1998 189–195
Kinetics of ultra-fast excited state proton transfer from
7-hydroxy-4-methylflavylium chloride to water
a
a
b
a,c,)
J.C. Lima , Isabel Abreu , Raymond Brouillard , Antonio L. Mac¸anita
´
a
Instituto de Tecnologia Quımica e Biologica, Apt. 127, 2780 Oeiras, Portugal
´
´
b
Laboratoire de Chimie des Polyphenols, associe au CNRS, Institut de Chimie, UniÕersite Louis Pasteur, Strasbourg, France
´
´
´
c
Instituto Superior Tecnico, AÕ. RoÕisco Pais, 1096 Lisboa codex, Portugal
´
Received 31 July 1998; in final form 19 October 1998
Abstract
Excited state proton transfer from 7-hydroxy-4-methylflavylium chloride to water is reported. From a modified analysis
Ž
.
of picosecond time-resolved fluorescence data not using the lifetime of a parent compound , all rate constants were
determined: the deprotonation rate constant of the flavylium cation, kds1.4=1011 sy1, the protonation rate constant of the
base form, kp s2.3=1010 l moly sy and the reciprocal fluorescence lifetimes of these species, kAH s7.8=10 sy
1
1
9
1
q
tAHqs128 ps and kAs7.6=10 sy1 tAs132 ps , in water, at 208C. The value of kd is the largest measured value for
9
Ž
.
Ž
.
Ž
.
an intermolecular proton transfer to water . q 1998 Elsevier Science B.V. All rights reserved.
q
Ž
. Ž . w
x
1. Introduction
Synthetic salts of 2-phenylbenzopyrilium fla-
cation AH
and the quinonoidal base A 11 .
With this simplification, the acid–base equilibrium
q
q
Ž
Ž
.
AH |AqH
can be analysed separately from
. w
x
vylium 1–8 have been shown to be of great useful-
ness for understanding the rather complex multiequi-
libria of the more heavily substituted natural
flavylium salts: the reddish or bluish anthocyanins,
the others which is particularly useful for studies in
the excited state.
Some of the multiequilibria of anthocyanins and
synthetic flavylium salts are strongly affected by
light absorption 1,12,13 , which has important con-
sequences on their photochemistryrphotodegra-
dation. Only in one case 14 have excited state rate
constants have been determined deprotonation and
back protonation of the cis-chalcone form of mal-
vidin 3,5-diglucoside .
w
x
found in most flowers and fruits 9,10 . As an exam-
ple, with 7-hydroxy-4-methylflavylium chloride
w
x
Ž
.
Ž
HMF only two species out of the five species in
w
x
multiequilibria, usually found with natural antho-
cyanins in aqueous solution at moderately acidic pH
Ž
. w x
values 9 are detected in acidic pH: the flavylium
.
Here, we report a full kinetic analysis of the
excited state proton exchange between the flavylium
cation of HMF and water, and the largest value, to
the best of our knowledge, found for a deprotonation
)
Corresponding author. Address: Instituto de Tecnologia
Quımica e Biologica, Departamento de Quimica do ISTLrUTL,
´
´
Apt. 127, Rua da Quinta Grande, 2780 Oeiras, Portugal. E-mail:
macanita@itqb.unl.pt
Ž
.
rate constant to water .
0009-2614r98r$ - see front matter q 1998 Elsevier Science B.V. All rights reserved.
Ž
.
PII: S0009-2614 98 01199-3

(
)
190
J.C. Lima et al.rChemical Physics Letters 298 1998 189–195
.
2. Experimental
prism polarizer . The emission passed through an
Ž
.
analyzing monochromator Jobin-Yvon H20 Vis and
HMF was prepared according to the published
was detected with a microchannel plate photomulti-
w
x
Ž
.
procedure 7,8 . The parent compound 4-methyl-7-
plier Hamamatsu R3809u-50 . The signal from the
photomultiplier was filtered in a constant fraction
Ž
.
methoxyflavylium chloride MMF was synthesised
Ž
.
from 4-methyl-7-methoxycoumarin and the appropri-
discriminator Canberra 2126 and used as stop sig-
Ž
.
Ž
ate Grignard reagent C6 H5MgBr in diethylether.
The solution turns yellow and the reaction was com-
plete after 4 h. The excess of Grignard reagent was
nal in the time-to-amplitude converter TAC Can-
berra 2145 . The signal from the TAC was sent to an
ADC Canberra 8213 and stored in a multichannel
card Canberra AccuSpec installed in a PC. Alter-
.
Ž
.
Ž
Ž
.
hydrolysed with an aqueous solution of HClO4 20%
3
.
Ž
vrv , the solid was filtered and dissolved in methanol
to precipitate the inorganic salts. The methanolic
nate measurements 10 counts at the maximum per
.
cycle of the excitation pulse profile and sample
emissions were made until a typical value of 5=103
solution was concentrated and purified by high-per-
Ž
.
formance liquid chromatography HPLC to separate
the excess of the unreacted coumarin. The eluent was
evaporated and the solid was dissolved in a 0.1 M
HCl solution and then precipitated by evaporation.
Twice distilled and deionized water was used for the
preparation of the solutions, and the pH values were
adjusted by addition of HCl and NaOH: for pH
values -2, analytical concentrations of Hq where
used and for pH values )2, the pH was measured at
208C with a Crison micropH 2002. UV–Vis absorp-
tion spectra were recorded on a Beckman DU-70
spectrophotometer. Fluorescence spectra were
recorded on a Spex Fluorolog F2121 and all spectra
where corrected for monochromator and photomulti-
plier response curves.
counts at maximum.
3. Results
3.1. Absorption spectra
Absorption spectra of HMF in aqueous solution,
at several pH values, are presented in Fig. 1. In the
pH range from 1.3 to 3.0, the flavylium cation
Ž
.
spectra 1 and 2 is essentially the sole species
present in solution. As the pH increases from 3.0 up
to 6.9, the characteristic flavylium cation AH
q
Ž
.
Ž
absorption band lmax s416.5 nm, ´s42900 l
moly1 cmy1 decreases, while the absorption band
.
Time-resolved fluorescence was measured using
the time-correlated single-photon counting technique
employing an excitation source consisting of a
Ž
mode-locked Ti:sapphire laser Spectra-Physics
.
Ž
.
Tsunami pumped by an argon ion Spectra-Physics
laser running on multiline. The pulse repetition rate
Ž
.
at the output of the Tsunami 82 MHz was reduced
Ž
to 800 kHz with an opto-acoustic modulator Spec-
tra-Physics Pulse Selector Model 3980 . The light
.
frequency of the Tsunami output was doubled using
Ž
.
a second-harmonic generator crystal LBO . The fun-
damental signal was monitored with a photodiode,
Ž
filtered in a constant fraction discriminator Canberra
.
2126 and used as start signal in the time-to-ampli-
Ž
.
tude converter TAC Canberra 2145 .
The samples where measured with excitation at
Fig. 1. Absorption spectra of 7-hydroxy-4-methylflavylium chlo-
w
Ž
413 nm the excitation light is depolarized Oriel
y5
Ž
.
M
ride ;5=10
in aqueous solutions as a function of pH:
.x
depolarizer ref. 28110 , and the emissions at 480
1spH 1.3; 2spH 2.2; 3spH 3.0; 4spH 4.8; 5spH 5.8;
6spH 6.1; 7spH 6.9. The inset of the figure represents the
molar fractions of AHq and A as a function of the pH.
and 600 nm were collected at 908 geometry with a
Ž
polarizer at 35.38 Spindler and Hoyer Glan-laser

(
)
J.C. Lima et al.rChemical Physics Letters 298 1998 189–195
191
observed. The isosbestic point at ls440 nm also
q
Ž
.
indicates that only the flavylium cation AH and
Ž .
the quinonoidal base A are in equilibrium within
this pH range.
3.2. Fluorescence emission and excitation spectra
Fluorescence emission spectra of HMF in water,
as a function of pH, are shown in Fig. 2. The
emission spectrum of the parent compound 7-me-
thoxy-4-methylflavylium chloride at pHs1 is also
Fig. 2. Emission spectra of 7-hydroxy-4-methylflavylium chloride
Ž
.
shown for comparison spectrum 5 . Excitation at
solution at: 1spH 7.2, 2spH 2.6, 3spH 1.2, 4spH 0.3;
Ž
.
415 nm flavylium cation absorption band , gives
Ž
.
lexc s415 nm. The spectrum at pH 7.2 line 1 was multiplied by
rise to two emission bands, one at lmax s484 nm
4 for comparison purposes. The fluorescence spectrum of the
Ž
.
very weak emission , and the other at lmax s610
Ž
.
parent compound line 5 at pH 1 at the same excitation wave-
length is also plotted.
Ž
.
nm strong emission , the last one with a shoulder at
580 nm. Between pHs0.3 and 2.6, where there is
no significant change in the flavylium cation concen-
y1
q
Ž
.
Ž
at lmax s464 nm ´s27700 l mol
cmy1 at-
tration AH is essentially the only species in the
Ž . Ž
.
.
tributed to the quinonoidal base A neutral form
ground state , the intensity of the 480 nm emission
Ž
.
increases Fig. 1, curves 3–5 . The value of the
pKa s4.4 is in agreement with the published values
band slightly decreases as the 610 nm emission
intensity increases.
The emission band at 610 nm matches the emis-
sion obtained by direct excitation of the quinonoidal
w
x
7,11 .
Other species, usually formed with natural antho-
. w
Ž
x
Ž
cyanins hemiacetals and chalcones 9,15 were not
base at pHs7.2 where A is the only species in the
Fig. 3. Fluorescence decays of 7-hydroxy 4-methylflavylium chloride in water at pH 1, 208C, excitation at 413 nm, collecting at 480 nm
. Ž . Ž .
Ž
.
Ž
flavylium emission and 600 nm base emission . a Global analysis; b deconvolution of the base decay with the flavylium cation decay.

(
)
192
J.C. Lima et al.rChemical Physics Letters 298 1998 189–195
.
ground state . Note that this emission is still ob-
Ž
.
served for very low pH values pH-1 , at which pH
the base is not present in the ground-state and only
the flavylium cation absorbs.
The fluorescence excitation spectrum from emis-
sion at lem s484 nm is identical to the absorption
spectrum of AHq, at any pH value, indicating that
the 484 nm band is due to the emission of that
q
Ž
.
species AH . Also, the excitation spectra at lem
s
610 nm matches the absorption of AHq for pH
values where this is the sole species in the ground
state. Therefore, the fluorescence emission band at
lmax s484 nm is assigned to the flavylium cation
q
Ž
.
AH and the emission band at lmax s610 nm is
Ž .
A form. The
assigned to the quinonoidal base
results are characteristic of an acid–base system
where the excited state pKa) is lower than the
w
x
ground state pKa, as found in phenols 16 and many
w
x
other molecules 16,17 .
3.3. Fluorescence decays
The fluorescence decays of HMF in water, with
Ž
excitation at 413 nm, at pHs1 only flavylium
.
cation present in the ground state , collected at 480
Ž .
Fig. 4.
flavylium cation and base decays, plotted as a function of H
Ž .
a
Decay times obtained from global analysis of the
Ž
.
Ž
.
nm flavylium emission and 600 nm base emission
w
q x
.
are shown in Fig. 3a . The decays can be fitted with
the sum of two exponentials with extremely short
lifetimes which are identical at the two emission
Ž
.
and b ratio of pre-exponential factors Rsa2 ra1 plotted as a
function of H
w
q x
.
Ž
Ž .
Ž .
.
and kA are the reciprocal fluorescence lifetimes of
AHq and A, respectively.
The time evolution of the concentrationrfluores-
cence of the two species obey the well-known equa-
wavelengths Eqs. 1 and 2 below . At 600 nm,
Ž
.
Ž
the shortest time t2 is a rise-time negative pre-ex-
ponential and the sum of the two pre-exponential
factors approaches zero, which means that the fluo-
rescence intensity of the base is zero at time zero.
When the concentration of Hq is decreased, the two
.
w
x
tions 18 :
2 t
IAHq t sa eyl t qa2 eyl
,
1
1
Ž .
Ž .
1
Ž
.
decay times do not appreciably change Fig. 4 a , but
2 t
IA t sa eyl t yeyl
,
2
Ž .
1
Ž
.
Ž .
Ž
.
the weight of the longest time t1 in the decay at
q
Ž
.
480 nm AH emission decreases and tends to zero
where l2 and l1 are the reciprocals of the shortest
at Hq concentrations -10y3 M the 480 nm decays
Ž
Ž
.
Ž .
t2 and the longest t1 decay times, respectively,
given by:
.
become single exponential . This is shown in Fig. 4b
Ž
where the ratio of pre-exponential factors Rs
2
qx
.
w
qx
a2ra1 is plotted as a function of H .
w
XqY" XyY q4kd kp H
(
Ž
.
l2, 1
s
,
3
Ž .
2
3.3.1. Kinetic analysis
with
The previous results are consistent with a simple
acid–base process in the excited state, as represented
in Scheme 1 where kd and kp are the deprotonation
and protonation rate constants, respectively, and kAHq
Xskd qkAHq
,
4
Ž .
w
qx
Yskp H qkA ,
5
Ž .

(
)
J.C. Lima et al.rChemical Physics Letters 298 1998 189–195
193
Scheme 1.
and the ratio of the pre-exponential factors Rsa2ra1
very small contribution of emission from the base at
Ž
is related to the kinetic constants by:
480 nm note in Fig. 2 the low fluorescence quantum
yield of AHq as compared to A would cause signif-
.
Xyl1
Ž
.
Rs
.
6
Ž .
icant changes to R R is very large . Second, the
l2 yX
choice of a model compound is generally problem-
w
x
atic 19 and, in particular here, the assumption of
identical photophysical properties for HMF and
MMF, in the strongly interacting solvent water, is
obviously dangerous.
Ž
An alternative to the above procedure not using
the pre-exponentials ratio or the fluorescence life-
time of the model compound is based on the evalua-
tion of Y by the deconvolution of the decay of the
.
Ž .
Ž
q. Ž
. w
Ž .
base A from the decay of the acid AH
Eq. 11
x
where the symbol m means convolution 20 .
IA t sk IAHq t m eyY t
.
11
Ž . Ž .
Ž
.
d
We have performed this kind of analysis and in
l2 Rql1
Rq1
Ysl2 ql1 yX ,
kd sXykAHq
and
Xs
,
7
Ž .
Ž
.
fact obtained perfect single exponential fits Fig. 3b .
Ž
.
The values of the decay times equal to 1rY are
clearly pH dependent Fig. 5a , in contrast with the
8
Ž .
9
Ž .
Ž
.
Ž
.
decay times obtained from global analysis Fig. 4a .
After determination of Y, the evaluation of the
four constants is straightforward. First, from the
qx Ž
linear plot of Y vs. H Fig. 5a the values of the
.
l yl 2 y XyY
2
w
Ž
.
Ž
.
2
1
9
y1
w
qx
s
kp H
.
10
Ž
.
Ž
.
Ž
tA s
two rate constants kA s 7.6"0.1 =10 s
4kd
132 ps and kp s 2.3"0.2 =10 l moly1 sy1 are
determined from the zero intercept and slope, respec-
10
.
Ž
.
There are two reasons to avoid this procedure for the
present case: first, the value of the pre-exponential
Ž
Ž ..
tively see Eq. 5 . The determination of the two
Ž .
remaining rate constants kd and kAHq is accom-
Ž .
factors ratio
R
can be uncertain because even a

(
)
194
J.C. Lima et al.rChemical Physics Letters 298 1998 189–195
Ž
Ž ..
its determination from kAHqsXykd Eq. 9 im-
possible. However, the plot of l1 ql2 vs. l1 l2
Fig. 5b allows a surprisingly accurate determina-
Ž
.
tion of kAHq due to cancellation of the errors arising
from the uncertainty of the values of the shortest
reciprocal decay time l2. Also, the value of the
y1
Ž
.
experimental zero intercept in Fig. 5b 11.4 ns
is
in excellent agreement with the one calculated using
y1
Ž
.
.
the values of kd, kA and kAHq 11.2 ns
Second, note that the value of the fluorescence
Ž
lifetime of the flavylium cation of HMF tAHqs128
.
Ž
ps is quite different from that of MMF tAHqs762
ps . This supports our concerns with respect to the
.
use of data obtained with the parent compound.
4.2. Comparison with steady-state fluorescence data
From the pH dependence of fluorescence intensi-
ties of AHq and A, the value of the proton transfer
yield h) and the value of the apparent acidity
constant in the excited state Ka)p have been evalu-
)
)
Ž
. w x
ated h s0.95 and Kap s5.0 21 . Both values are
Ž
.
consistent with those calculated with Eqs. 14 and
)
Ž
.
15 and time-resolved fluorescence: h s0.93 and
Ka)p s4.9.
h) sk r k qk
Ž .
w
q x
Ž .
Fig. 5. a Plot of Y as a function of H and b plot of l1ql2
vs. l1 l2.
AHq
,
14
15
Ž
.
Ž
Ž
.
.
d
d
K ) s k rk
rkA .
.
AHq
k
Ž
.
Ž
plished using the values of l1 and l2 after appropri-
ap
d
p
Ž . Ž . Ž
Ž .
ate re-arrangements of Eqs. 3 – 5 Eqs. 12 and
Ž ..
13 .
4.3. Rate constants
kA
1
l1 ql2 skd 1y
qkAHq
q
l1 l2 ,
ž
/
kAHq
kAHq
The value of the deprotonation constant of HMF
to water kd s1.4=10 sy1 is extremely large to
the best of our knowledge, it is the largest value
measured for an intermolecular proton transfer to
11
Ž
.
Ž
12
13
Ž
Ž
.
.
kd sl1 ql2 yYykAHq
.
.
water . Although this value does not exclude a diffu-
From the slope of the linear plot of l1 ql2
vs. l1 l2 Fig. 5b , kAHqs 7.8"0.1 =10 sy1
9
Žw
x
.
sional mechanism H2O s55 M , preliminary re-
sults obtained in methanolic solutions in presence of
smaller concentrations of water indicate the presence
of a static mechanism involving the hydrogen-bonded
Ž
.
Ž
.
Ž
.
Ž .
tAHqs128 ps is obtained and from Eq. 12 ,
11
kd s 1.4"0.3 =10 sy1
.
Ž
.
w
x
to water HMF molecule 22 .
Ž
The value of the protonation rate constant kp s
4. Discussion
2.3=1010 l moly1 sy1 is within the diffusion-con-
.
trol region but it may contain some proton migration
contribution. Both problems are presently being ad-
dressed from the influence of water concentration
and temperature on the fluorescence decays.
4.1. Data analysis
y1
Ž
.
First, note that the value of kAyHq1 7.8 ns
is
.
which makes
Ž
lower than the error in kd "30 ns

(
)
J.C. Lima et al.rChemical Physics Letters 298 1998 189–195
195
w x
2
F. Pina, L. Benedito, M.J. Melo, A.J. Parola, M.A. Bernardo,
5. Conclusions
Ž
.
J. Chem. Soc., Faraday Trans. 92 1996 1693.
w x
3
F. Pina, M.J. Melo, R. Ballardini, L. Flamigni, M. Maestri,
The excited state proton transfer reactions be-
tween HMF and water are very fast, leading to short
decay times with the inherent measurement difficul-
ties. Neither the parent compound lifetime nor the
pre-exponentials ratio can be used for data analysis.
The determination of all rate constants in Scheme 1
is possible by the use of non-standard methods of
analysis of time-resolved fluorescence data. These
are applicable under particular conditions, such as
this one, where the kinetics depends on the proton
concentration.
Ž
.
New J. Chem. 21 1997 969.
w x
4
F. Pina, M.J. Melo, M. Maestri, R. Ballardini, V. Balzani, J.
Ž
.
Am. Chem. Soc. 119 1997 5556.
w x
5
Ž
.
.
.
R.A. McClelland, S. Gedge, J. Am. Chem. Soc. 102 1980
5838.
w x
6
Ž
R.A. McClelland, G.H. McGall, J. Org. Chem. 47 1982
3730.
w x
7
Ž
G. Mazza, R. Brouillard, J. Agric. Food Chem. 35 1987
422.
w x
8
R. Brouillard, G.A. Iacobucci, J.G. Sweeny, J. Am. Chem.
Ž
.
Soc. 104 1982 7585.
w x
9
Ž
.
R. Brouillard, in: J.B. Harborne Ed. , The Flavonoids, Ad-
vances in Research, Chapman and Hall, London, 1988, p.
525.
w
x
Ž
.
10 R. Brouillard, in: P. Markakis Ed. , Anthocyanins as Food
Colors, ch. 1, Academic Press, New York, 1982.
Acknowledgements
w
w
x
Ž
.
11 J.M. Baranac, D. Amic, J. Serb. Chem. Soc. 50 1985 299.
´
x
12 R. Matsushima, H. Mizuno, A. Kajiura, Bull. Chem. Soc.
Ž
.
Jpn. 67 1994 1762.
This work was partially supported by the research
grant PRAXISrPCEXrCrQUIr56r96 and the IC-
CTIrCAPESr423. Partial funding by the CIENCIA
Program for the acquisition of the picosecond laser
system is also acknowledged. JCL and IA are grate-
ful for the PRAXIS 4r4.1rBPDr3410 and PRAXIS
4r4.1rBDr3589 grants, respectively. ALM thanks
George Striker for the newest version of his decon-
volution program and K.A. Zachariasse for intercali-
bration measurements of fluorescence decays.
w
w
w
w
w
w
w
w
w
w
x
13 R. Matsushima, H. Mizuno, H. Itoh, J. Photochem. Photo-
Ž
.
biol. A: Chem. 89 1995 251.
x
14 J.C. Lima, P. Danesh, P. Figueiredo, F.S. Pina, A.L.
Ž
.
Mac¸anita, Photochem. Photobiol. 59 1994 412.
x
15 H. Santos, D.L. Turner, J.C. Lima, P. Figueiredo, F. Pina,
Ž
.
A.L. Mac¸anita, Phytochemistry 33 1993 1227.
x
16 R.S. Becker, Theory and Interpretation of Fluorescence and
Phosphorescence, Wiley-Interscience, New York, 1969.
x
17 L.G. Arnaud, S.J. Formosinho, J. Photochem. Photobiol. A:
Ž
.
Chem. 75 1993 1.
x
18 J.B. Birks, Photophysics of Aromatic Molecule, Wiley, Lon-
don, 1970.
x
19 A.L. Mac¸anita, J. Magalhaes, A. Dias, H. Teles, E. Iglesias,
˜
´
Ž
.
J. Chem. Soc., Faraday Trans. 86 1990 4011.
x
Ž
.
20 J.C. Conte, J.M.G. Martinho, Chem. Phys. Lett. 134 1987
References
350.
x
21 F. Pina, M.J. Melo, H. Santos, J.C. Lima, I. Abreu, R.
w x
1
P. Figueiredo, J.C. Lima, H. Santos, M.C. Wigand, R.
Brouillard, A.L. Mac¸anita, F. Pina, J. Am. Chem. Soc. 116
1994 1249.
Ž
.
Ballaldini, M. Maestri, New J. Chem. 1998, in press .
x
22 J.C. Lima, Ph.D. Thesis, Instituto Superior Tecnico, Lisboa,
´
Ž
.
1996, p. 87.