Excited-state interactions in diastereomeric flurbiprofen-thymine dyads

Article abstract of DOI:10.1021/jp3063838

Excited-state interactions between (S)- or (R)-flurbiprofen ((S)- or (R)-FBP) and thymidine (dThd) covalently linked in dyads 1 or 2 have been investigated. In both dyads, the only emitting species is ¹FBP, but with a lower fluorescence quantum yield (φ_F) and a shorter fluorescence lifetime (τ_F) than when free in solution. These results indicate that dynamic quenching occurs either by electron transfer or via exciplex formation, with FBP as the charge-donating species. In acetonitrile, both mechanisms are favored, while in dioxane exciplex formation is predominating. Isomer 1 displays lower values of φ_F and τ_F than its analogue 2, indicating that the relative spatial arrangement of the chromophores plays a significant role. The triplet quantum yields (φ_T) of 1 and 2 are significantly higher than the expectations based solely on ¹FBP-dThd intersystem crossing quantum yields (φ_ISC), with φ_T (1) > φ_T (2). This can be explained in terms of intramolecular charge recombination at the radical ion pairs and/or the exciplexes, which would be again dependent on geometrical factors. The triplet lifetimes (τ_T) of ³FBP-dThd and free ³FBP* are similar, indicating the lack of excited-state interactions at this stage. The FBP-dThd dyads could, in principle, constitute appropriate model systems for the elucidation of the excited-state interactions in noncovalent DNA-ligand complexes.

Full text of DOI:10.1021/jp3063838

Article
pubs.acs.org/JPCA
Excited-State Interactions in Diastereomeric Flurbiprofen−Thymine
Dyads
Paula Bonancía,† Ignacio Vaya,† M. Jose Climent,† Thomas Gustavsson,‡ Dimitra Markovitsi,‡
́
́
M. Consuelo Jimenez,*,† and Miguel A. Miranda*,†
́
†Chemistry Department/Institute of Chemical Technology UPV-CSIC, Universitat Politec
̀ ̀
nica de Valencia, 46022 Valencia, Spain
‡CNRS, IRAMIS, SPAM, Francis Perrin Laboratory, URA 2453, 91191 Gif-sur-Yvette, France
S
* Supporting Information
ABSTRACT: Excited-state interactions between (S)- or (R)-
flurbiprofen ((S)- or (R)-FBP) and thymidine (dThd)
covalently linked in dyads 1 or 2 have been investigated. In
1
both dyads, the only emitting species is FBP*, but with a
lower fluorescence quantum yield (ϕF) and a shorter
fluorescence lifetime (τF) than when free in solution. These
results indicate that dynamic quenching occurs either by
electron transfer or via exciplex formation, with FBP as the
charge-donating species. In acetonitrile, both mechanisms are
favored, while in dioxane exciplex formation is predominating.
Isomer 1 displays lower values of ϕF and τF than its analogue 2, indicating that the relative spatial arrangement of the
chromophores plays a significant role. The triplet quantum yields (ϕT) of 1 and 2 are significantly higher than the expectations
based solely on 1FBP*−dThd intersystem crossing quantum yields (ϕISC), with ϕT (1) > ϕT (2). This can be explained in terms
of intramolecular charge recombination at the radical ion pairs and/or the exciplexes, which would be again dependent on
3
3
geometrical factors. The triplet lifetimes (τT) of FBP*−dThd and free FBP* are similar, indicating the lack of excited-state
interactions at this stage. The FBP−dThd dyads could, in principle, constitute appropriate model systems for the elucidation of
the excited-state interactions in noncovalent DNA−ligand complexes.
are usually marketed as racemic mixtures (with few exceptions,
such as naproxen). However, their anti-inflammatory activity is
mainly attributed to the (S)-enantiomer.14 The lowest singlet
and triplet excited states of flurbiprofen (FBP) have recently
been characterized,15 and their photophysical properties have
been employed to obtain relevant information on different
issues, either from a biological point of view (binding degree to
a transport protein, enzyme-like activity of HSA, etc.)16−20 or
for analytical purposes (i.e., determination of enantiomeric
compositions).21 In addition, FBP has been used as a triplet
sensitizer of quinolones, for the determination of the triplet
energy of thymine in DNA.22,23
Model systems (dyads) containing a photoactive drug
covalently linked to a biomolecule fragment have been designed
in the past with the purpose to provide mechanistic insight into
the key basic processes such as energy transfer, electron
transfer, or exciplex formation.24−27 Following this strategy, we
have now synthesized two stereoioisomeric dyads 1 and 2
(Chart 1) containing (S)- or (R)-FBP attached to thymidine at
the 5′ position of the sugar. These dyads are, a priori, suitable
models for the investigation of excited-state interactions in their
complexes with DNA.
INTRODUCTION
■
The fact that ultraviolet (UV) radiation induces chemical
reactions in DNA leading to carcinogenic mutations is at the
origin of numerous investigations devoted to the study of the
electronically excited states of DNA.1−5 The UV solar radiation
that reaches the Earth’s surface ranges from 400 to 320 nm
(UVA) and from 320 to 290 nm (UVB). Both UVA and UVB
radiations are able to induce damage in DNA, which is
responsible for some types of cancer.6−9 Concerning UVB, it is
efficiently absorbed by nucleobases, causing direct photo-
reactions of DNA. It is also possible to cause DNA degradation
by means of photosensitization, performed either by
endogenous (i.e., metabolites) or by exogenous (i.e., drugs or
cosmetics) agents.10 It is known that nonsteroidal anti-
inflammatory drugs (NSAIDs), when interacting with free
nucleobases or nucleosides, enhance the photoinduced
formation of thymine cyclobutane dimers.11,12 In this context,
a detailed understanding of deactivation channels of drug
excited states in the presence of biomolecules is essential to
predict a possible photobiological risk in patients, mainly
associated with phototoxicity.
Among NSAIDs, which are widely prescribed therapeutic
agents for a variety of pathological conditions including
headache, pain caused by degenerative diseases, and fever,13
2-arylpropionic acid derivatives are an important group. They
include ibuprofen, naproxen, ketoprofen, or flurbiprofen and
Received: June 28, 2012
Revised: August 7, 2012
Published: August 8, 2012
© 2012 American Chemical Society
8807
dx.doi.org/10.1021/jp3063838 | J. Phys. Chem. A 2012, 116, 8807−8814

The Journal of Physical Chemistry A
Article
removed under reduced pressure and the viscous residue
dissolved in dichloromethane (20 mL) and washed with
NaHCO3 (1 M), to remove unreacted FBP and dThd. The
organic phase was dried with anhydrous sodium sulfate, and the
solvent was rotavaporated. The organic residue was submitted
to column chromatography using hexane/methanol (9:1, v/v)
Chart 1. Chemical Structures of the Dyads Investigated in
This Work
1
as eluent, to produce 1 or 2 (ca. 70%). Copies of the H and
13C NMR spectra are shown in Figures S1 and S2 of the
Supporting Information.
5′-(S)-[(2-Fluoro-1,1′-biphenyl-4-yl)propanoyl]-
thymidine (1). White powder. UV λmax (log ε): 250 (4.1).
FTIR ν (cm−1): 3420 (br), 1721, 1655, 1480, 1418, 1267, 1174,
767, 697. 1H NMR (δ, ppm): 1.43 (d, 3H, J = 7.2 Hz), 1.72 (d,
3H, J = 1.2 Hz), 1.81−1.90 (m, 1H), 2.09−2.17 (m, 1H), 3.80
(q, 1H, J = 7.1 Hz), 3.97−4.01 (m, 1H), 4.16−4.21 (m, 2H),
4.32 (dd, 1H, J = 12.0, J = 6.0 Hz), 6.08 (dd, 1H, J = 8.1, J = 6.0
Hz), 7.02−7.11 (m, 2H), 7.23−7.41 (m, 7H). 13C NMR (δ,
ppm): 11.5, 17.6, 39.5, 45.0, 64.4, 71.4, 85.0, 85.5, 110.3, 114.9
+ 115.2, 123.6, 127.6 + 123.7, 127.6, 128.0 + 128.1, 128.3,
128.6, 128.7, 130.9, 135.4, 136.1, 142.3 + 142.4, 150.9, 158.1 +
161.4, 165.1, 173.9. HRMS for C25H26FN2O6 (MH+): calcd,
469.1775; found, 469.1773.
Steady-state and time-resolved spectroscopic measurements
were performed on the dyads to characterize the excited-state
properties. To correctly interpret the results obtained on the
dyads, comparative measurements were also done on the
individual chromophores (FBP and dThd) as well as on their
equimolar mixture.
The results show that singlet excited-state interactions
between the FBP and dThd chromophores are only observed
when both units are covalently linked. The only emitting
species is 1FBP*, which undergoes dynamic quenching by
either electron transfer or exciplex formation, with FBP as the
charge-donating species. Because both mechanisms require
interaction at a contact distance, the relative spatial arrange-
ment of the chromophores plays a significant role. Intra-
molecular charge recombination at the radical ion pairs and/or
the exciplexes results in population of the FBP triplet excited
state, a pathway that is also sensitive to geometrical factors. In
noncovalent FBP/DNA complexes, excitation in the UVC
could result in charge injection from 1FBP* to a nearby
thymine, and the excess electron could be transported along the
DNA double strand.28
5′-(R)-[(2-Fluoro-1,1′-biphenyl-4-yl)propanoyl]-
thymidine (2). White powder. UV λmax (log ε): 250 (4.1).
FTIR ν (cm−1): 3422 (br), 1720, 1656, 1475, 1418, 1272, 1172,
1
768, 698. H NMR (δ, ppm): 1.43 (d, 3H, J = 7.2 Hz), 1.55−
1.64 (m, 1H), 1.81 (d, 3H, J = 1.2 Hz), 1.97−2.05 (m, 1H),
3.77 (q, 1H, J = 7.2 Hz), 3.96−3.99 (m, 1H), 4.05−4.09 (m,
1H), 4.14 (dd, 1H, J = 12.2, J = 3.1 Hz), 4.39 (dd, 1H, J = 12.2,
J = 5.1 Hz), 6.06 (dd, 1H, J = 8.1, J = 6.0 Hz), 7.00−7.08 (m,
2H), 7.23−7.40 (m, 7H). 13C NMR (δ, ppm): 11.3, 17.4, 39.3,
44.8, 64.5, 71.2, 84.6, 85.6, 110.4, 114.8 + 115.1, 123.6 + 123.7,
127.6, 127.9 + 128.1, 128.3, 128.6 + 128.7, 130.8, 135.4, 136.2,
142.1 + 142.2, 151.0, 158.1 + 161.4, 165.1, 173.9. HRMS for
C25H26FN2O6 (MH+): calcd, 469.1775; found, 469.1766.
Fluorescence Measurements. Emission spectra were
recorded on a spectrofluorometer system, which was provided
with a monochromator in the wavelength range of 200−900
nm. The solutions were placed into 10 mm × 10 mm quartz
cells. The absorbance of the samples at the excitation
wavelength was kept below 0.2. Steady-state experiments
were performed at 22 °C or at −78 °C. For the fluorescence
quantum yield determination, FBP in MeCN/air was used as a
secondary reference, with ϕF = 0.17 (air) or 0.21 (N2).25
Time-resolved fluorescence measurements were performed
using the fluorescence upconversion (FU) and time-correlated
single photon counting (TCSPC) techniques. For both
techniques, the excitation source was the third harmonic (267
nm) of a mode-locked Ti-Sapphire laser, delivering 100 fs
pulses with a repetition rate of 76 and 4.75 MHz, respectively
(using a pulse-picker). The femtosecond FU setup has been
described earlier.29 For the upconversion technique, the average
excitation power used was 40 mW. The power density cannot
be measured precisely within the excitation volume, but we
estimate it to 0.2 0.1 GW/cm2 for a 40 mW output from the
tripler unit (assuming a 40 μm diameter of the focused beam).
The fwhm value of a Gaussian apparatus function is about 350
fs. We judge that the time resolution of the setup is better than
100 fs after deconvolution, depending on the signal-to-noise
ratio. Solutions (about 25 mL) were kept flowing through a 0.4
mm quartz cell, which itself was kept in continuous motion
perpendicular to the excitation beam to minimize thermal
EXPERIMENTAL SECTION
■
Materials and Solvents. The (S)- and (R)-flurbiprofen
((S)- and (R)-FBP), thymidine (dThd), and 1-(3-dimethyla-
minopropyl)-N-ethylcarbodiimide hydrochloride (EDC) were
commercially available. Their purity was checked by 1H nuclear
magnetic resonance (NMR) and high-performance liquid
chromatography (HPLC) analysis. Spectroscopic-grade sol-
vents were used without further purification.
General. Absorption spectra were recorded using 1 cm
quartz cells; λmax and log ε, in brackets, are given for the main
absorption bands. Infrared (IR) spectra were obtained with a
Fourier transform infrared (FTIR) instrument; νmax (cm−1) is
1
given for the main absorption bands. H NMR spectra were
measured in CD3OD, using a 300 MHz apparatus; chemical
shifts are reported in δ (ppm) values, using tetramethylsilane
(TMS) as an internal standard. High-resolution mass
spectrometry (HRMS) was performed by ESI+ with a
XevoQToF spectrometer, in the ITQ, at Universitat Politec
̀
nica
de Valencia.
̀
Synthesis of the Substrates. The FBP−dThd dyads 1 and
2 (Chart 1) were obtained by condensation of (S)- or (R)-
flurbiprofen and thymidine at the 5′ position of the sugar as
follows: to a mixture of dThd (0.30 g, 1.23 mmol) and (S)- or
(R)-FBP (0.29 g, 1.19 mmol) in 2 mL of anhydrous pyridine
was slowly added EDC (0.23 g, 1.19 mmol) under magnetic
stirring, at 0 °C. When the addition was completed, the
reaction mixture was kept at 0 °C for 2 h. Pyridine was then
8808
dx.doi.org/10.1021/jp3063838 | J. Phys. Chem. A 2012, 116, 8807−8814

The Journal of Physical Chemistry A
Article
Figure 1. (A) UV absorption spectra in MeCN of (S)-FBP (10−5 M), dThd (10−5 M), and a 1:1 molar mixture of (S)-FBP and dThd at the same
concentration, together with that calculated for the same mixture taking into account independent absorption by each chromophore. (B)
Fluorescence spectra of isoabsorptive solutions (A267 = 0.1) of (S)-FBP (black) and (S)-FBP/dThd at 1:1 molar ratio (red), in acetonitrile. The gray
trace corresponds to the theoretical band calculated as independent emission of FBP and dThd, taking into account the photons absorbed by each
chromophore.
effects. All measurements were performed at room temperature
(22 1 °C) under aerated conditions.
sampling). The monitoring beam is provided by a ceramic
xenon lamp and delivered through fiber optic cables. The laser
pulse is probed by a fiber that synchronizes the LFP system
with the digitizer operating in the pretrigger mode. All transient
spectra were recorded using 10 mm × 10 mm quartz cells with
a capacity of 4 mL and were bubbled for 30 min with N2 before
signal acquisition. The absorbance of the samples was 0.2 at the
laser excitation wavelength (266 nm).
For the TCSPC technique, a Becker & Hickl GmbH PC card
was used. A Schott WG 295 long pass filter was placed in front
of a SPEX monochromator. The detector was a microchannel
plate (R1564 U Hamamatsu) providing an instrumental
response function of 60 ps (fwhm). The average laser power
(0.1 mW) was measured with a Melles Griot broadband
powermeter. The irradiated area on the surface of the cell was
ca. 0.2 cm2 corresponding to a pulse intensity of 2.4 kW/cm2.
Solutions were contained in a 10 mm × 10 mm quartz cell and
continuously stirred. Successive recordings with the same
sample gave identical decays, which were eventually merged to
improve the signal-to-noise ratio. Such a procedure allowed us
to ensure that the measured signals were not altered during the
measurements due to a possible accumulation of photo-
products.
For both FU and TCSPC, parallel (Ipar(t)) and perpendicular
(Iperp(t)) excitation/detection measurements were performed
by controlling the polarization of the exciting beam with a zero-
order half-wave plate. The transmission under parallel and
perpendicular conditions was found to be identical, so no
correction was applied. From these measurements, the
fluorescence anisotropy can be calculated from the formula
r(t) = (Ipar(t) − Iperp(t))/(Ipar(t) + 2Iperp(t)).
Phosphorescence Measurements. Emission spectra were
recorded on a spectrofluorometer system, which was provided
with a monochromator in the wavelength range of 200−900
nm. The solutions were placed into a quartz tube and
introduced into liquid nitrogen prior the measurements. The
absorbance of the samples was 0.3 at the excitation wavelength
(266 nm).
Laser Flash Photolysis Experiments. Laser flash
photolysis (LFP) experiments were performed by using a Q-
switched Nd:YAG laser (Quantel Brilliant, 266 nm, 10 mJ per
pulse, 5 ns fwhm) coupled to a mLFP-111 Luzchem
miniaturized equipment. This transient absorption spectrom-
eter includes a ceramic xenon light source, 125 mm
monochromator, Tektronix 9-bit digitizer TDS-3000 series
with 300 MHz bandwidth, compact photomultiplier and power
supply, cell holder and fiber optic connectors, fiber optic sensor
for laser-sensing pretrigger signal, computer interfaces, and a
software package developed in the LabVIEW environment from
National Instruments. The LFP equipment supplies 5 V trigger
pulses with programmable frequency and delay. The rise time
of the detector/digitizer is ∼3 ns up to 300 MHz (2.5 GHz
RESULTS AND DISCUSSION
■
Absorption and Fluorescence Properties of an
Equimolar FBP/dThd Mixture. The absorption and emission
spectra of (S)-FBP are well-known15 and do not deserve a
detailed description here. Concerning dThd, only spectra in
aqueous solution have been reported.30 The absorption and
emission spectra of dThd in MeCN and dioxane do not differ
substantially in shape from those in water and are given in the
Supporting Information (Figures S4 and S5). It should be
remarked that the molar extinction of dThd is significantly
higher in MeCN than in water (15.3 × 103 at 265 nm as
compared to 9.9 × 103 at 267 nm). Likewise, the fluorescence
quantum yield of dThd in MeCN is substantially lower than
that in water (3.0 × 10−5 as compared to 1.5 × 10−4).
The fluorescence spectrum of the equimolar mixture in
MeCN (under aerated conditions) with λexc = 267 nm is shown
in Figure 1B. It is nearly identical to that of (S)-FBP obtained
under the same conditions after scaling with the relative
absorbance (48%), calculated using isoabsorptive solutions of
(S)-FBP and dThd as models for comparison. Indeed, the
fluorescence quantum yield (ϕF) of dThd (3 × 10−5, see Figure
S5 in the Supporting Information) is several orders or
magnitude lower than that of (S)-FBP (0.17). This shows
that the emission of the equimolar mixture of (S)-FBP and
dThd is only due to the photons absorbed by (S)-FBP, also
indicating the absence of any excited-state interaction in the
mixture (Figure 1A).
Time-resolved fluorescence measurements, both on the pico-
and on the nanosecond time scales, were done on dThd and
FBP as well as their mixture in MeCN (22 °C, λexc = 267 nm).
Comparative measurements were made at 310 nm; this
wavelength was chosen because it corresponds to the maximum
of the FBP fluorescence spectrum.
The fluorescence decay of dThd in aqueous solution is
known to be ultrafast and hardly depends on the emission
wavelength, with an average lifetime well below 1 ps.29 This is
due to the presence of highly efficient intramolecular
8809
dx.doi.org/10.1021/jp3063838 | J. Phys. Chem. A 2012, 116, 8807−8814

The Journal of Physical Chemistry A
Article
nonradiative deactivation channels. The fluorescence decay of
dThd in MeCN recorded by FU is likewise ultrafast and can be
described by a biexponential function with time constants of
0.25 and 1.65 ps (see below). The fluorescence of FBP in
MeCN is much longer-lived, in line with the much higher
quantum yield. Interestingly, the decay kinetics depends on the
emission wavelength (see Figure S3).
The fluorescence trace of the mixture at 310 nm rises
instantaneously before undergoing an ultrafast decay (20%
amplitude, limited by the instrumental apparatus function) to a
constant value. This is in line with the superposition of dThd
and FBP fluorescence scaled by the relative absorbance at 267
nm and emission intensities at 310 nm. Time-resolved data thus
show no indication of any excited-state interaction in the
mixture.
Neither steady-state nor time-resolved data thus show
indications of FBP−dThd interaction in the mixture.
Photophysical Behavior of FBP/dThd Dyads. The UV
absorption spectra of dyads 1 and 2 are nearly identical to the
sum of the spectra of the individual FBP and dThd units
obtained under the same conditions (Figure 2). This shows the
absence of any significant ground-state intramolecular inter-
actions between the two chromophores in the two dyads.
nm)15 rather than to dThd for the reasons evoked above
(Figure 3A). The fluorescence quantum yields (ϕF) of the two
dyads were somewhat higher in dioxane than in acetonitrile
(see Table 1).
In ethanol matrix at −78 °C, the fluorescence intensity shows
the same trend as the relative absorbance of FBP in the dyads
(Figure 3B). This value is much higher than that observed at 22
°C, showing the near absence of the fluorescence quenching.
This may be the result of the limited degrees of freedom for
conformational relaxation in the rigid matrix that prevents
effective population of the conformations required for intra-
molecular excited-state interactions.
Fluorescence decays on the femtosecond time scale were
recorded for the dyads at 310 nm by FU at 22 °C, λexc = 267
nm (Figure 4). In MeCN solution, the kinetics were very
similar to that of the equimolar mixture, that is, characterized
by an instantaneous rise followed by an ultrafast decay leveling
out to a constant value (∼80%, see Figure 4A). The ultrafast
component is typical for the <1 ps fluorescence decay of dThd,
as in the case of the equimolar mixture. The subsequent
constant level emission is supposedly only due to 1FBP*. These
observations clearly demonstrate two different phenomena: (i)
the lack of intramolecular singlet−singlet energy transfer from
1dThd* to FBP and (ii) the FBP (biphenyl) singlet state as the
only species responsible for the residual emission observed in
the steady-state fluorescence. No differences were found on this
short time scale between the two dyads.
However, on the nanosecond time scale (TCSPC), both
dyads showed a more complex behavior, either in acetonitrile
or in dioxane (Figure 4B). While the fluorescence decay of (S)-
FBP is monoexponential, that of the dyads is clearly
1
nonexponential and much faster, indicating a dynamic FBP*
quenching. Interestingly, the (S)- isomer decays faster than the
(R)-derivative. To quantify this comparison, experimental data
were fitted using a multiexponential model function convoluted
by the apparatus function. As expected, the (S)-FBP
fluorescence profiles are well described by one single time
constant (1.70 ns in MeCN, 1.80 ns in dioxane). Conversely, a
triexponential model function was required for the dyads, with
a ∼100 ps, a ∼300 ps, and a low-amplitude nanosecond
component close to the FBP lifetime. The average lifetimes
(⟨τ⟩ = a1τ1 + a2τ2) of the (S)- and (R)-isomers in MeCN are
0.27 and 0.31 ns, respectively (Table 1), but all three time
constants for the (S)-isomer are faster than those of the (R)-
isomer. In addition to the low-amplitude nanosecond
component, the fluorescence decays below 1 ns can be
correctly described only by a biexponential model function.
The characteristic times obtained are τ1 = 0.14 ns (a1 = 0.50)
Figure 2. UV absorption spectra of isoabsorptive solutions (at 267
nm) of (S)-FBP (black), dThd (dark green), 1 (red), and 2 (blue), in
acetonitrile.
The fluorescence spectra of the two dyads as well as (S)-FBP
obtained at room temperature after excitation at 267 nm are
shown in Figure 3A. The ordinate in this figure is chosen as a
relative fluorescence intensity, that is, scaled for equal
absorbance at the excitation wavelength (the spectra are
normalized with regards to the maximum of the FBP spectrum
taken equal to one). The most remarkable feature observed is a
dramatic fluorescence quenching in the two dyads. The
observed fluorescece was assigned to (S)-FBP (λem= 310
Figure 3. Fluorescence spectra of isoabsorptive solutions at λexc = 267 nm of (S)-FBP (black), 1 (red), and 2 (blue). (A) At 22 °C, in acetonitrile
(dashed line) and dioxane (solid line). (B) In ethanol matrix at −78 °C.
8810
dx.doi.org/10.1021/jp3063838 | J. Phys. Chem. A 2012, 116, 8807−8814

The Journal of Physical Chemistry A
Article
a
Table 1. Photophysical Properties of (S)-FBP and the Dyads at λexc = 267 nm
b
c
ϕF
τF (ns)
kF (×10−8 s−1
)
kQ (×10−9 s−1
Acetonitrile
)
ϕQ
ϕisc
ϕT
ϕT−ϕisc
(S)-FBP
0.17
1.70
0.27
0.31
1.00
1.16
1.21
0.71
0.71
1
2
0.015
0.018
3.11
2.64
0.402
0.393
0.063
0.075
0.167
0.126
0.104
0.051
Dioxane
(S)-FBP
0.24
1.80
0.51
0.62
1.33
1.49
1.54
0.70
0.70
1
2
0.035
0.044
1.40
0.329
0.303
0.102
0.127
0.219
0.167
0.117
0.040
1.06
a
b
In general, errors were lower than the last significant figure of the stated values. Measured at 310 nm; for the dyads, the fitting was performed using
c
a triexponential model function, and the values given here correspond to an average of the three lifetimes. In deaerated media, λexc = 266 nm.
Figure 4. Fluorescence decays at 310 nm of (S)-FBP (black), dThd (green), 1 (red), and 2 (blue) in acetonitrile/air, at 22 °C, using 267 nm as the
excitation wavelength. (A) Picosecond time scale (FU), in acetonitrile. (B) Nanosecond time scale (TCSPC), in acetonitrile and dioxane.
and τ2 = 0.47 ns (a2 = 0.50) for (S)-FBP and τ1 = 0.18 ns (a1 =
0.84) and τ2 = 0.70 ns (a2 = 0.16) for (R)-FBP. The need for
two fast components may indicate the presence of two types of
conformations with different efficiencies of the quenching
mechanism (vide infra). It is worth mentioning that whatever
the nature of this quenching mechanism, it is more efficient in
the case of the (S)-isomer. A similar trend was observed in
dioxane, although with markedly slower time constants.
rotational time of about 44 ps, very close to that experimentally
observed (45 ps, r0 = 0.31). The monoexponential behavior,
and in particular the lack of any fast decay at early times, shows
that there is no internal rotation between the two residues.
Upon laser flash photolysis (LFP) of the two dyads
(excitation at 266 nm), the typical FBP triplet−triplet
absorption spectrum was obtained (Figure 6A). The signals
were less intense than when FBP was directly excited under the
same conditions, taking into account its relative absorbance in
the dyads. Interestingly, the intensity was higher than what was
expected in view of the dramatic fluorescence decrease (vide
supra). This indicates that the singlet excited state is not the
only precursor of the triplet state. Moreover, the triplet
quantum yield (Table 1) was found to be higher for the (S)-
than for the (R)-stereoisomer. Quantitative data determined
from these experiments are shown in Table 1.
The fluorescence anisotropy decays (recorded by FU) of the
dyads (Figure 5) were compared to that of free FBP. For the
The triplet lifetimes in the dyads are equal to that of free FBP
(24 μs),15 indicating the lack of intramolecular interactions in
the excited triplet state.
In contrast, the phosphorescence spectra of the dyads
obtained in a frozen ethanol matrix at −78 °C (Figure 6B)
revealed that the emission from the triplet excited state merely
reflects the fraction of photons absorbed by the FBP
chromophore.
Figure 5. Fluorescence anisotropies (at 22 °C) of (S)-FBP (black), 1
(red), and 2 (blue) in acetonitrile/air at 1.0 × 10−3 M concentration,
using 267 nm as the excitation wavelength.
The fluorescence quantum yields of the dyads (ϕF dyad) were
obtained using (S)-FBP as standard (ϕFstd = 0.21 in MeCN/N2
or 0.17 in MeCN, air).15 In this expression (eq 1), Adyad and
Astd are the areas under the emission curve for dyads and FBP,
respectively, Abs represents the absorbance at the excitation
wavelength, and η corresponds to the refractive index of the
solvent. In all cases, the fluorescence was quenched by more
than 80%. Stereodifferentiation was observed in the fluo-
rescence lifetimes, which were much shorter for the dyads than
for (S)-FBP, indicating a dynamic quenching. From these
experimentally determined ϕF and τF values, and taking into
latter, the anisotropy decayed with a characteristic time of about
30 ps, while for the dyads the decay time was about 45 ps. For
both FBP and the dyads, an initial fluorescence anisotropy at
time zero r0 = 0.32 was observed. FBP has a molecular volume
of ca. 300 Å3, as calculated from the van der Waals
increments.31 Using the Stokes−Einstein−Debye equation,32
one thus obtain a rotational time of about 27 ps, slightly faster
than that observed (30 ps). The deviation can be understood in
terms of the nonspherical shape of the chromophore. The total
volume of the FBP−dThd dyads is about 492 Å3, giving a
8811
dx.doi.org/10.1021/jp3063838 | J. Phys. Chem. A 2012, 116, 8807−8814

The Journal of Physical Chemistry A
Article
Figure 6. (A) Laser flash photolysis spectra (λexc = 266 nm) of (S)-FBP (black), 1 (red), and 2 (blue), in acetonitrile or dioxane, at 1.0 × 10−5
M
concentration, 1 μs after the laser pulse. (B) Phosphorescence spectra (λexc = 266 nm) of (S)-FBP (black), 1 (red), and 2 (blue) in ethanol matrix at
−78 °C.
Table 2. Free Energy Changes (in kcal mol−1) Associated with Electron Transfer or Exciplex Formation in FBP*/dThd
a
Systems
acetonitrile
dioxane
ΔGET
ΔGexc
ΔGET
ΔGexc
donor/acceptor
S1
T1
S1
T1
S1
T1
S1
T1
FBP/dThd
dThd/FBP
−30.0
−2.4
+4.0
−25.1
+2.5
+8.9
−5.0
+22.7
+29.0
+56.7
−22.7
+4.9
+11.2
+38.9
+31.6
+36.5
a
The given results are precise to 10% because the exact values of the redox potentials are not known for all compounds.
account the fraction of photons absorbed by (S)-FBP (Ψ = 0.48
in acetonitrile and 0.46 in dioxane), the kF value was obtained
using eq 2. The subsequent quenching constants (kQdyad) and
quenching quantum yields (ϕQdyad) were determined by means
of eqs 3 and 4, respectively, where the average lifetimes τFdyad
are used for convenience, even if the best fitting is
nonmonoexponential.
Likewise (eq 5), it was possible to determine the intersystem
crossing quantum yield (ϕISCdyad). The triplet quantum yields
(ϕTdyad) were determined from LFP experiments (eq 6) at λexc
= 266 nm, where ΔODdyad and ΔODstd are the top of the
optical density value for dyads and FBP (the standard),
respectively, at the maximum of the T−T absorption spectrum
(360 nm). The molar absorption coefficient (ε) was assumed to
be the same in the given solvent.
Concerning the nature of the 1FBP* fluorescence quenching
mechanism, it can be due to energy transfer, electron transfer,
or exciplex formation. The possibility of Forster energy transfer
̈
can be reasonably disregarded, in view of the sensitivity of ϕQ
to the rigidity of the medium (no quenching in solid matrix)
and to the polarity of the solvent (lower quenching efficiency in
dioxane than in acetonitrile).33 Moreover, the observed
stereodifferentiation in the ϕF and ϕT values also supports
that a close contact between the two chromophores is necessary
in the involved processes.
As regards the feasibility of formation of radical ion pairs or
exciplexes from the singlet or triplet states, it can be estimated
by application of eqs 7 and 8:34
ΔGET(kcal mol−1
)
⎡
⎤
2.6
ε
donor
OX
acceptor
RED
ηd2yad
= 23.06 E
− E
+
− 0.13 − E*
Adyad
⎢
⎣
⎥
⎦
Absstd
(7)
ϕF = ϕF
η2
dyad
std
Astd Absdyad
(1)
std
⎡
ΔGexc(kcal mol−1) = 23.06 Edonor − Eacceptor
⎢
OX
RED
⎣
ϕF /ψ
dyad
kF
=
⎤
⎛
⎜
⎞
ε − 1
2ε − 1
dyad
τF
⎟
− 0.75
− 0.19 + 0.38
⎥
(2)
(3)
dyad
⎝
⎠
⎦
1
τF
1
(8)
− E*
kQ
=
−
dyad
dyad
τF
Taking into account the corresponding redox potentials
(EOX(biphenyl) = 1.95 V,35 ERED(thymidine) = −1.10 V,36
EOX(thymidine) = 1.70 V,37 and ERED(biphenyl) = −2.55 V38),
dyad
std
kQ
kF
dyad
the appropriate dielectric constants (εMeCN = 37.5 and εdioxane
=
ϕQ
=
× ϕF
dyad
2.25),39 and the FBP singlet or triplet energies (E*),15 the free
energy values obtained for electron transfer and exciplex
formation are shown in Table 2.
(4)
(5)
(6)
dyad
ϕ
ISCstd
ϕISC = ϕF
×
In general, charge transfer is only favored from the FBP
singlet excited state as donor, and the fluorescence quenching
may be due both to electron transfer and to exciplex formation.
The latter process should be clearly predominating in dioxane.
dyad
dyad
ϕF
std
ΔOD
ΔOD εT
εT
dyad
std
3
ϕT = ϕT
By contrast, FBP* should be essentially unreactive in both
dyad
std
std
dyad
solvents, which is in agreement with the lack of dynamic
8812
dx.doi.org/10.1021/jp3063838 | J. Phys. Chem. A 2012, 116, 8807−8814

The Journal of Physical Chemistry A
Article
quenching observed in the LFP experiments. Moreover, the
positive values of ΔG consistently obtained in Table 2 for the
T1 columns indicated that charge recombination leading to
3FBP* is indeed thermodynamically allowed.
CONCLUSIONS
■
Singlet excited-state interactions between the FBP and dThd
chromophores are only observed when both units are
covalently linked. There is no indication of any intermolecular
interactions in an equimolar mixture, neither in the ground
states, nor in the excited states. In the FBP−dThd dyads, the
Another point of interest is the origin of the FBP triplet. Its
quantum yield (ϕT), determined from LFP, is found to be
significantly higher than ϕISC. Triplet generation could in
principle occur through several pathways: (a) intersystem
crossing from S1, (b) triplet−triplet energy transfer from
3dThd*, or (c) charge recombination of the radical ion pairs or
exciplexes possibly resulting from quenching of the excited
singlet states. As ϕT is higher than ϕISC, path (a) cannot be the
only operating route. Contribution of path (b) is possible
because ET(dThd) is ca. 9 kcal mol−1 higher than ET(FBP);
however, on the basis of the known ϕISC values for dThd (in
the order of 10−2), this process would not account for the
observed ϕT, at least for 1. Hence, pathway (c) also seems to be
operating, and its contribution is clearly more important for the
(S)-isomer than for its (R)-analogue.
1
only emitting species is FBP*, with a quantum yield and a
lifetime remarkably lower than when acting as a free
chromophore. Free energy changes, calculated using Rehm−
Weller equations, indicate that dynamic quenching may occur
either by electron transfer or by exciplex formation, with FBP as
the charge-donating species. In acetonitrile, both mechanisms
are clearly favored, while in dioxane exciplex formation should
be overwhelmingly predominating. In all cases, quenching
requires interaction at contact distance, so the relative spatial
arrangement of the chromophores plays a significant role.
Indeed, the (S)-isomer (1) exhibits lower values of ϕF and τT
than does its (R)-analogue. As a general observation, the triplet
quantum yield (ϕT) determined in the dyads is significantly
higher than the expectations based on the 1FBP*−dThd
intersystem crossing quantum yields (ϕISC) taking into account
the fraction of light absorbed by the FBP moiety. This can be
accounted for in terms of intramolecular charge recombination
at the radical ion pairs and/or the exciplexes. Again, these
processes would be expected to depend on geometrical factors,
and indeed their contribution is more important in the case of 1
According to the results obtained from the photophysical
measurements and the calculations described above, the main
deactivation processes that can take place upon excitation of a
dyad containing covalently linked FBP and dThd subunits are
indicated in Scheme 1 and can be summarized as follows. Initial
Scheme 1. Photophysical Processes in FBP−dThd Dyads
3
than in 2. Once formed, the FBP*−dThd dyads have triplet
lifetimes (τT) similar to that of free FBP, indicating the lack of
excited-state interactions at this stage. This is in agreement with
thermodynamic calculations, which lead to positive ΔG values
for electron transfer or exciplex formation occurring from the
triplet excited state.
In principle, the FBP−dThd dyads could be appropriate
models to anticipate the nature of excited-state interactions in
noncovalent FBP/DNA complexes. Thus, upon excitation in
the UVC, the resulting 1FBP* would inject negative charge to a
nearby thymine, and the excess electron could travel along the
DNA double strand, reaching a distant base or eventually
repairing a bipyrimidine lesion.42−45
ASSOCIATED CONTENT
■
S
* Supporting Information
Figures S1 and S2: 1H NMR and 13C NMR spectra of 1 and 2.
Figure S3: Fluorescence decays of FBP and FBP+dThd in
MeCN (20 °C, λexc = 267 nm) at different emission
wavelengths. Figure S4: Steady-state absorption spectra of
dThd in MeCN and water. Figure S5: Corrected steady-state
fluorescence spectra of dThd in MeCN and thymine in water
used for the calculation of the fluorescence quantum yields.
Figure S6: Model structures of 1 and 2 obtained using
HyperChem Release 8.0.3 for Windows Molecular Model
System. This material is available free of charge via the Internet
at http://pubs.acs.org.
excitation at 267 nm (i) leads to the first singlet excited state of
either FBP or dThd, as expected from the UV absorption
spectra of these chromophores (Figure 1). Singlet−singlet
energy transfer (SSET) from 1dThd* (Es ca. 100 kcal mol−1)37
to the FBP moiety (Es ca. 99 kcal mol−1)15 would be nearly
thermoneutral, but such a process is kinetically ruled out, in
view of the ultrafast decay traces observed in Figure 4A.
Deactivation of 1dThd* occurs through a highly efficient
intramolecular nonradiative decay40 (ii) or by intersystem
crossing as a minor pathway41 (iii). Triplet−triplet energy
transfer from 3dThd* (74 kcal mol−1)22,23 to FBP (ET = 65 kcal
mol−1)15 is outlined in (iv). Fluorescence emission from 1FBP*
is represented in (v), and intramolecular quenching (Q) of
1FBP* to afford radical ion pairs or exciplexes is illustrated by
process (vi). Route (vii) corresponds to intersystem crossing
(ISC) in FBP. Finally, charge recombination (ChR) at the
AUTHOR INFORMATION
■
Corresponding Author
*E-mail: mmiranda@qim.upv.es.
3
Notes
radical ion pairs or the exciplexes leading also to FBP* is
indicated in (viii).
The authors declare no competing financial interest.
8813
dx.doi.org/10.1021/jp3063838 | J. Phys. Chem. A 2012, 116, 8807−8814

The Journal of Physical Chemistry A
Article
(30) Onidas, D.; Markovitsi, D.; Marguet, S.; Sharonov, A.;
Gustavsson, T. J. Phys. Chem. B 2002, 106, 11367−11374.
(31) Edward, J. T. J. Chem. Educ. 1970, 47, 261−270.
(32) Lakowicz, J. R. Principles of Fluorescence Spectroscopy, 3rd ed.;
Springer-Verlag: New York, 2006.
ACKNOWLEDGMENTS
■
Financial support from the Spanish Government (Grants
CTQ2010-14882 and CTQ2009-13699), Salvador Madariaga
Program (Grant to M.C.J.), Juan de la Cierva Program
(contract to I.V.), Generalitat Valenciana (Prometeo 2008/
090), and the Universitat Politecnica de Valencia (PAID 05-11,
̀ ̀
ref 2766) is gratefully acknowledged.
(33) Speiser, S. Chem. Rev. 1996, 96, 1953−1976.
(34) Weller, A. Z. Phys. Chem. 1982, 133, 93−98.
(35) Merkel, P. B.; Luo, P.; Dinnocenzo, J. P.; Farid, S. J. Org. Chem.
2009, 74, 5163−5173.
(36) Steenken, S.; Telo, J. P.; Novais, H. M.; Candeias, L. P. J. Am.
Chem. Soc. 1992, 114, 4101−4709.
REFERENCES
■
(1) Crespo-Hernan
Chem. Rev. 2004, 104, 1977−2020.
(2) Middleton, C. T.; de La Harpe, K.; Su, C.; Law, Y. K.; Crespo-
́
dez, C. E.; Cohen, B.; Hare, P. M.; Kohler, B.
(37) Vicic, D. A.; Odom, D. T.; Nunez, M. E.; Gianolio, D. A.;
́
̃
McLaughlin, L. W.; Barton, J. K. J. Am. Chem. Soc. 2000, 122, 8603−
8611.
́
Hernandez, C. E.; Kohler, B. Annu. Rev. Phys. Chem. 2009, 60, 217−
(38) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of
Photochemistry, 2nd ed.; Marcel Dekker: New York, 1993.
(39) Lide, D. R. Handbook of Chemistry and Physics, 85th ed.; CRC
Press: Boca Raton, FL, 2004.
(40) Gustavsson, T.; Banyasz, A.; Lazzarotto, E.; Markovitsi, D.;
Scalmani, G.; Frisch, M. J.; Barone, V.; Improta, R. J. Am. Chem. Soc.
2006, 128, 607−619.
239.
(3) Gustavsson, T.; Improta, R.; Markovitsi, D. J. Phys. Chem. Lett.
2010, 1, 2025−2030.
(4) Markovitsi, D.; Gustavsson, T.; Vaya,
1, 3271−3276.
(5) Cuquerella, M. C.; Lhiaubet-Vallet, V.; Bosca,
Chem. Sci. 2011, 2, 1219−1232.
́
I. J. Phys. Chem. Lett. 2010,
́
F.; Miranda, M. A.
(41) Serrano-Perez, J. J.; Gonzal
́ ́
ez-Luque, R.; Merchan, M.; Serrano-
(6) Osterlind, A. Pharmacol. Toxicol. 2009, 72, 67−68.
(7) Soehnge, H.; Ouhtit, A.; Ananthaswamy, H. N. Front. Biosci.
1997, 2, D538−D551.
Andres
́
, L. J. Phys. Chem. B 2007, 111, 11880−11883.
(42) Fazio, D.; Trindler, C.; Heil, K.; Chatgilialoglu, C.; Carell, T.
Chem.-Eur. J. 2011, 17, 206−212.
(8) Mouret, S.; Baudouin, C.; Charveron, M.; Favier, A.; Cadet, J.;
Douki, T. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 13765−13770.
(43) Park, M. J.; Fujitsuka, M.; Kawai, K.; Majima, T. Chem.-Eur. J.
2012, 18, 2056−2062.
́
(9) Bayansz, A.; Vaya, I.; Changenet-Barret, P.; Gustavsson, T.;
(44) Breeger, S.; Hennecke, U.; Carell, T. J. Am. Chem. Soc. 2004,
126, 1302−1303.
Douki, T.; Markovitsi, D. J. Am. Chem. Soc. 2011, 133, 5163−5165.
(10) Cadet, J.; Douki, T.; Ravanat, J.-L.; Di Mascio, P. Photochem.
Photobiol. Sci. 2009, 8, 903−911.
(45) Giese, B.; Carl, B.; Carl, T.; Carell, T.; Behrens, C.; Hennecke,
U.; Schiemann, O.; Feresin, E. Angew. Chem., Int. Ed. 2004, 43, 1848−
1851.
(11) Chouini-Lalanne, N.; Defais, M.; Paillous, N. Biochem.
Pharmacol. 1998, 55, 441−446.
(12) Marguery, M. C.; Chouini-Lalanne, N.; Ader, J. C.; Paillous, N.
Photochem. Photobiol. 1998, 68, 679−684.
(13) Brooks, P. M.; Day, R. O. N. Engl. J. Med. 1991, 324, 1716−
1725.
(14) Evans, A. M. J. Clin. Pharmacol. 1996, 36, 7S.
(15) Jimen
Photobiol. Sci. 2004, 3, 1038−1041.
(16) Vaya, I.; Bueno, C. J.; Jimen
ChemMedChem 2006, 1, 1015−1020.
(17) Jimenez, M. C.; Miranda, M. A.; Vaya,
127, 10134−10135.
(18) Vaya, I.; Jimen
112, 2694−2699.
(19) Bueno, C. J.; Jimen
2009, 113, 6861−6865.
(20) Perez-Ruiz, R.; Bueno, C. J.; Jimen
Phys. Chem. Lett. 2010, 1, 829−833.
(21) Vaya, I.; Bueno, C. J.; Jimenez, M. C.; Miranda, M. A. Chem.-
Eur. J. 2008, 14, 11284−11287.
(22) Bosca, F.; Lhiaubet-Vallet, V.; Cuquerella, M. C.; Castell, J. V.;
́ ́
ez, M. C.; Miranda, M. A.; Tormos, R.; Vaya, I. Photochem.
́
́
ez, M. C.; Miranda, M. A.
́
́
I. J. Am. Chem. Soc. 2005,
́
́
ez, M. C.; Miranda, M. A. J. Phys. Chem. B 2008,
́
ez, M. C.; Miranda, M. A. J. Phys. Chem. B
́
́
ez, M. C.; Miranda, M. A. J.
́
́
́
Miranda, M. A. J. Am. Chem. Soc. 2006, 128, 6318−6319.
́
(23) Lhiaubet-Vallet, V.; Cuquerella, M. C.; Castell, J. V.; Bosca, F.;
Miranda, M. A. J. Phys. Chem. B 2007, 111, 7409−7414.
́ ́
(24) Jimenez, M. C.; Stiriba, S. E.; Tormos, R.; Perez-Prieto, J.;
Miranda, M. A. Photochem. Photobiol. Sci. 2004, 3, 36−38.
(25) Vaya, I.; Jimenez, M. C.; Miranda, M. A. J. Phys. Chem. B 2007,
111, 9363−9371.
(26) Jimenez, M. C.; Pischel, U.; Miranda, M. A. J. Photochem.
Photobiol., C 2007, 8, 128−142.
(27) Vaya, I.; Perez-Ruiz, R.; Lhiaubet-Vallet, V.; Jimen
Miranda, M. A. Chem. Phys. Lett. 2010, 486, 147−153.
́
́
́
́
́
́
ez, M. C.;
(28) Wagenknecht, H.-A. Principles and Mechanisms of Photo-
induced Charge Injection, Transport and Trapping in DNA. In Charge
Transfer in DNA: From Mechanism to Application; Wagenknecht, H.-A.,
Ed.; Wiley-VCH: Weinheim, 2005; pp 1−26.
(29) Gustavsson, T.; Sharonov, A.; Markovitsi, D. Chem. Phys. Lett.
2002, 351, 195−200.
8814
dx.doi.org/10.1021/jp3063838 | J. Phys. Chem. A 2012, 116, 8807−8814