Time-resolved spectroscopic study of the photochemistry of tiaprofenic acid in a neutral phosphate buffered aqueous solution from femtoseconds to final products

Article abstract of DOI:10.1021/jp310315f

The photo-decarboxylation and overall reaction mechanism of tiaprofenic acid (TPA) was investigated by femtosecond transient absorption (fs-TA), nanosecond transient absorption (ns-TA), and nanosecond time-resolved resonance Raman (ns-TR³) spec

Full text of DOI:10.1021/jp310315f

Article
pubs.acs.org/JPCB
Time-Resolved Spectroscopic Study of the Photochemistry of
Tiaprofenic Acid in a Neutral Phosphate Buffered Aqueous Solution
from Femtoseconds to Final Products
Tao Su, Jiani Ma, Ming-De Li, Xiangguo Guan, Lihong Yu, and David Lee Phillips*
Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong, P.R. China
S
* Supporting Information
ABSTRACT: The photo-decarboxylation and overall reaction
mechanism of tiaprofenic acid (TPA) was investigated by
femtosecond transient absorption (fs-TA), nanosecond transient
absorption (ns-TA), and nanosecond time-resolved resonance
Raman (ns-TR3) spectroscopic experiments in a neutral
phosphate buffered solution (PBS). In addition, density func-
tional theory (DFT) calculations were presented to help interpret
the experimental results. Resonance Raman and DFT calculation
results revealed that the deprotonated tiaprofenic acid (TPA−)
form was the primary species that is photoexcited in a near
neutral PBS aqueous solution. The fs-TA experimental data indicated that the lowest lying excited singlet state S1 underwent an
efficient intersystem crossing process (ISC) to quickly transform into the lowest lying excited triplet state T1 that then undergoes
decarboxylation to generate a triplet biradical species (TB3). ns-TA and ns-TR3 results observed a protonation process for TB3 to
produce a neutral species (TBP3) that then decayed via ISC to produce a singlet TBP species that further reacted to make the
final product (DTPA). A comparison of the present results for TPA− with similar results for the deprotonated form of ketoprofen
(KP−) in the literature was done to investigate how the thiophene moiety in TPA− that replaces one phenyl ring in KP− affects
the reaction mechanism and photochemistry of these nonsteroidal anti-inflammatory drugs (NSAIDs).
produce the singlet form (TB). TB3 and TB are both in
equilibrium with their conjugate acids. Subsequent protonation
of TB leads to the formation of the major photoproduct, 2-
INTRODUCTION
■
Nonsteroidal anti-inflammatory drugs (NSAIDs) are widely
utilized for the treatment of inflammation and inflammatory
diseases even though these types of drugs may cause various
adverse side effects such as cutaneous phototoxic responses in
benzoyl-5-ethylthiophene (DTPA), in aqueous solutions (see
Scheme 1 for an overview of the proposed mechanism).3
some patients.1,2 Tiaprofenic acid, 2-(5-[2-benzoyl] thienyl)
The photo-decarboxylation of TPA was also investigated by
density functional theory (DFT) calculations, and another
propionic acid (TPA), is a typical NSAID drug that can act as a
photosensitizer against cell components.3,4 TPA is usually
mechanism involving excitation and photo-decarboxylation of
the deprotonated form was proposed.24 This work concluded
dispensed and used as a racemic mixture for the treatment of
acute and chronic arthritis and osteoarthritis.5 Several
that TPA is not capable of undergoing the photo-decarbox-
ylation reaction from any of the lowest lying excited singlet
investigations have been reported for the photochemistry of
TPA and some analogues.6−17 One series of studies proposed
states, since the energy curves for those processes are all
that the photo-decarboxylation for TPA− involved the (n, π*)
endothermic or are associated with very high barriers. However,
triplet state of TPA− as a key intermediate and the
once the system has undergone efficient ISC to a
corresponding triplet state, the photo-decarboxylation reaction
decarboxylation occurred in a manner similar to those observed
can occur spontaneously. The decarboxylated product of TPA
for some other NSAIDs that contain a benzophenone (BP)
group.2,18−21 The triplet (n, π*) was proposed to have a higher
energy and to exist in thermal equilibrium with the lowest lying
triplet (π, π*) state. Upon absorption of a photon, TPA reaches
an excited singlet state and quickly undergoes an intersystem
has several options for further reaction that can lead to the
formation of the closed-shell ethyl derivative, alcohol or ketone
species, or a peroxyl radical capable of initiating lipid
peroxidation reactions. In several of these reactions, either
superoxide anions or singlet oxygen can be formed. It was
concluded that the most likely pathway involves the
protonation of the terminal (decarboxylated) carbon, followed
by decay from the triplet to the corresponding singlet ground
crossing (ISC) process to the lowest lying excited triplet (π,
π*) state (denoted as TPA3-1) with an approximately 90%
efficiency. The transformation to an upper excited triplet (n,
π*) state (TPA3-2) by thermal activation of an about 7−10 kJ/
mol energy barrier was proposed to lead to a decarboxylation
reaction that has an about 25% efficiency13,22,23 to form a triplet
Received: October 18, 2012
Published: December 11, 2012
biradical species (TB3) which subsequently undergoes ISC to
© 2012 American Chemical Society
811
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
Scheme 1. Proposed Photo-Decarboxylation Reaction Pathway of TPA in PBS Solution by Some Previous Experimental Studies
Scheme 2. Proposed Photo-Decarboxylation Reaction Pathways of TPA from Previous DFT Computational Results24
state. Protonation of the ketone connecting the two rings is also
possible, but this leads to the generation of a compound with a
long lifetime that requires a proton shift for further decay (see
Scheme 2 for a summary of the proposed mechanisms).
In the studies done so far for TPA, only a few experimental
Figure 1. The chemical structures of TPA and KP.
techniques were used to investigate the transient species during
the course of the photo-decarboxylation reaction of TPA in
aqueous solutions. Laser flash photolysis (LFP) played an
important role for the initial assignment of the transient species
involved in the photo-decarboxylation process of TPA.
However, it is sometimes difficult to clearly distinguish various
transient species that have similar electronic absorption spectra
and time scales in transient absorption experiments.20,22 In
these cases, it is sometimes useful to employ time-resolved
vibrational spectroscopy techniques that can more clearly
distinguish the various transient species present by their
vibrational frequencies that contain structural information and
can serve as a fingerprint for each species. Recently, we have
utilized ns-TR3 spectroscopy to carry out a number of studies
on the photochemistry of BP and NSAID drugs like ketoprofen
(KP) that contain a benzophenone group. These studies
provided new insight into the identities, structures, and
properties of several transient species associated with photo-
chemical reactions observed in these benzophenone containing
molecular systems.25−32 While both TPA and BP containing
NSAID drugs like KP both contain an aromatic carbonyl
moiety, the carboxyl containing side chain is attached to a
different ring group in these two compounds, as shown in
Figure 1. The carboxyl containing side chain is attached to a
phenyl group in KP, while it is attached to a thiophene group in
TPA. Here, we have used femtosecond transient absorption
spectroscopy (fs-TA), nanosecond transient absorption spec-
troscopy (ns-TA), and time-resolved resonance Raman (TR3)
spectroscopy to obtain valuable information about the
identities, structures, and properties of the transient species
from femtoseconds to the formation of the final product
DPTA. In order to mimic the physiological environment, the
predominantly PBS mixed solvent of 70% PBS/30% acetonitrile
(by volume ratio) was used due to the low solubility of TPA in
PBS. The results obtained for TPA here are compared and
contrasted with the photo-decarboxylation reactions of KP so
as to learn more about how the use of a thiophene versus a
phenyl ring connecting the carbonyl group to the carboxyl
containing side chain affects the photophysical and photo-
chemical processes in these NSAID drugs.
EXPERIMENTAL AND COMPUTATIONAL
■
METHODS
Materials. TPA was commercially obtained from Tractus
(with >96% purity) and used as received. Unless noted
otherwise, all of the sample solutions of TPA were used in a
812
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
mixture of PBS and MeCN with a volume ratio of 7:3. The
solutions used in the resonance Raman and ns-TR3 experiments
had a concentration of 1.0 mM. The pH 7.4 phosphate buffer
solution was prepared using 29.38 g of Na2HPO4·12H2O and
2.48 g of KH2PO4/L to give a 0.1 M solution.
Steady-State UV/vis Absorption. Steady-state absorption
spectra were recorded on a PerkinElmer Lambda 19 UV/vis
spectrometer using a 1 cm path length quartz cuvette.
decarboxylation. The energy for the pump and probe pulses
was in the range 2.5−3.5 mJ with a 10 Hz repetition rate. The
two Nd:YAG lasers were synchronized electronically by a pulse
delay generator to control the time delay of the pump and
probe lasers with the time delay between the laser pulses
monitored by a fast photodiode and 500 MHz oscilloscope, and
the time resolution for the ns-TR3 experiments was
approximately 10 ns. The pump and probe laser beams were
lightly focused onto the flowing sampling system, and the
Raman light was collected using reflective optics into a
spectrometer whose grating dispersed the light onto a liquid
nitrogen cooled CCD detector. The Raman signal was
accumulated for 30 s by the CCD before reading out to an
interfaced PC computer, and 10−20 scans of the signal were
added together to get a resonance Raman spectrum. The ns-
TR3 spectra presented here were obtained by the subtraction of
a resonance Raman spectrum with a negative time delay of
−100 ns (probe-before-pump spectrum) from the resonance
Raman spectrum with a positive time delay (pump−probe
spectrum), and the Raman shifts were calibrated by the known
MeCN solvent Raman bands with an estimated accuracy of 5
cm−1.
Density Functional Theory (DFT) Computations. DFT
calculations were done employing the Becke three-parameter
hybrid method with the Lee−Yang−Parr correlation functional
approximation (U)B3LYP method with a 6-311G(d,p) basis
set. The Raman spectra were obtained using a determination of
the Raman intensities from transition polarizabilities calculated
by numerical differentiation, with an assumed zero excitation
frequency. A Lorentzian function with a 15 cm−1 bandwidth for
the vibrational frequencies and an individual frequency scaling
factor (see Table2) was used in the comparison of the
fs-TA Experiments. The fs-TA measurements were
performed on the basis of a femtosecond Ti:sapphire
regenerative amplified Ti:sapphire laser system and an
automated data acquisition system. The amplifier was seeded
with the 120 fs output from the oscillator. The probe pulse was
obtained by using approximately 5% of the amplified 800 nm
output from the amplifier to generate a white-light continuum
(320−800 nm) through a CaF2 crystal. The maximum extent of
the temporal delay was about 3300 ps for the optical stage used
in the experiments, and the instrument response function was
determined to be about 150 fs. At each temporal delay, the data
were averaged for 2 s. The probe beam was split into two
before passing through the sample with one beam traveling
through the sample and the other one passing directly through
the reference spectrometer that monitors the fluctuations in the
probe beam intensity. Fiber optics were coupled to a
multichannel spectrometer with a CMOS sensor that had a
1.5 nm intrinsic resolution. For the present experiments, the
sample solutions were excited by a 267 nm pump beam (the
third harmonic of the fundamental 800 nm light from the
regenerative amplifier). Sample solutions of 60 mL were
employed in a flowing 2 mm path-length cuvette with a sample
UV/vis intensity of about 1 throughout the data acquisition.
The data were stored as three-dimensional (3D) wavelength−
time−absorbance matrices that were exported for use with the
fitting software.
ns-TA Experiments. Nanosecond time-resolved transient
absorption (ns-TA) measurements were carried out with a
LP920 laser flash spectrometer provided by Edinburgh
Instruments Ltd. The probe light source was a 450 W ozone
free Xe arc lamp with 10 Hz to single shot operation in a
sample chamber with an integral controller, high speed pump
and probe port shutters, sample holder, and filter holders. The
lamp produces a continuous spectrum between 150 and 2600
nm. Measurements of the ns-TA spectra were performed
according to the following procedure. Fresh sample solutions
were excited by a Q-switched Nd:YAG laser (fourth harmonic
line at λ = 266 nm). The probe light from a pulsed xenon arc
lamp passed through various optical elements, samples, and a
monochromator before being detected by a fast photomultiplier
tube and recorded with a TDS 3012C digital signal analyzer. In
the kinetics mode, a photomultiplier detector or InGaAs PIN
detector was used and the transient signal was acquired using a
fast, high resolution oscilloscope. In the spectral mode, an array
detector was fitted to the spectrograph exit port to measure a
full range of wavelengths simultaneously. About 4 mL solutions
were employed in a 10 mm path-length cuvette with a sample
of UV/vis intensity of about 1 throughout the data acquisition.
ns-TR3 Experiments. The ns-TR3 experiments were done
using an experimental apparatus and methods discussed in
detail previously,33,34 so only a brief description will be given
here. The 266 nm pump laser pulse and a 341.5 nm probe laser
pulse (generated from the second Stokes hydrogen Raman shift
laser line from the fourth harmonic) were used to detect
transient species and the final product arising from photo-
Table 1. Absorption Band(s) and Assignments of the
Transient Species Observed in the fs-TA Spectra (See Text
for Details)
transient species
absorption peaks (nm)
1, Sn
340
2, S1
360
3, T1
4, TB3
360, 600
360, 420, 600
Table 2. The Different Scaling Factors Used for the Normal
Raman Spectra to Compare to the Individual Transient
Species ns-TR3 Spectra
transient species
scaling factors
TPA−
0.961
TB3
TBP3
0.99
TBP
DTPA
0.969
0.983
0.974
calculated results with the experimental spectra. No imaginary
frequency modes were observed at the stationary states of the
optimized structures, and only one imaginary frequency was
observed for the saddle point transition state structures. The
isomerization process was explored by optimizing the structures
of the reactant, transition states, and product complexes.
Transition states were located using the Berny algorithm.
Frequency calculations at the same level of theory were
performed to confirm that the structures were at local minima
with all-real frequencies or at transition states with only one
imaginary frequency. The nature of the transition states was
determined by analyzing the motion by the eigenvector
associated with the imaginary frequency. Intrinsic reaction
813
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
coordinates (IRC)35−37 were calculated for the transition states
to confirm the relevant structures connect the two relevant
minima. The polarizable continuum model (PCM) is used for
evaluating the (bulk) solvent effects.38 No imaginary frequency
modes were observed at the stationary states of the optimized
structures. All the calculations were done using the Gaussian 03
program39 installed on the High Performance Computing
cluster at the Computer Centre in The University of Hong
Kong.
Previous studies found that the photo-decarboxylation
reaction takes place from the deprotonated form of TPA in a
PBS aqueous solution;20,22 however, no explicit structural
characterization of the properties of the ground electronic state
of the substrate species being excited were presented. In this
work, resonance Raman experiments were performed in a basic
aqueous solution at pH 13 and a neutral PBS aqueous solution.
Since the pKa of TPA is reported40 to be 3.0, the deprotonated
form of TPA (denoted as TPA−) is expected to be the main
species present in a pH 13 aqueous solution. The resonance
Raman spectrum obtained in a pH 13 aqueous solution is
compared with the predicted normal Raman spectrum
computed at the B3LYP/6-311G(d,p) level of theory in Figure
3. Inspection of Figure 3 shows there is good agreement
RESULTS AND DISCUSSION
■
UV/vis Absorption and Resonance Raman Spectra for
TPA in Neutral Aqueous PBS Solution. The UV/vis
experiments were conducted in both pure acetonitrile
(MeCN) and neutral PBS aqueous solutions, as shown in
Figure 2. The absorption spectrum of TPA (in its dissociated
Figure 3. Comparison of (a) the experimental resonance Raman
spectrum of TPA obtained in a basic aqueous solution with pH 13
with (b) the calculated ground state normal Raman spectrum of TPA−.
Figure 2. UV/vis absorption spectra of TPA in pure MeCN and
neutral aqueous PBS solvents.
between the experimental resonance Raman spectrum with the
predicted normal Raman spectrum, and this indicates that the
TPA− species is the main form present in the pH 13 aqueous
solution. Most of the resonance Raman bands observed for the
TPA− species are due to vibrations associated with the ring C
C stretching, CC stretching, and CO stretching motions
located in the 800−1800 cm−1 region. For instance, the CC
stretching and CO stretching modes are attributed to the
Raman bands at 1637 and 1663 cm−1. The ring breathing
vibrational modes are associated with the Raman bands at 1016
and 1156 cm−1. More detailed information about the
comparison of the experimental and calculated Raman spectra
and the assignments of the vibrational motions is listed in Table
7S of the Supporting Information.
The resonance Raman spectrum for TPA in a neutral PBS
aqueous solution was also obtained and compared with the
resonance Raman spectrum acquired in the pH 13 aqueous
solution, as shown in Figure 4. Examination of Figure 4 reveals
that the spectra are nearly identical, which indicates that the
species in the neutral PBS aqueous solution is also the TPA−
form of the molecule.
form) in the neutral PBS aqueous solution is characterized by
intense bands with maxima at 314 and 261 nm and a weak tail
extending up to 380 nm. The solvent polarity affects the
spectrum to a slight extent so that the two main bands at 314
and 261 nm shift to 300 and 259 nm, respectively, in pure
MeCN which has a lower solvent polarity. In contrast, the
absorption at λ > 360 nm does not experience any significant
solvent shift.
TD-DFT calculations were done to simulate the absorption
spectrum (see Figure 1S and Table 2S in the Supporting
Information). Frontier orbitals were also calculated at the
B3LYP/6-311G(d,p) level of theory and indicate that both of
the two main bands are characteristic of (π, π*) absorption
transitions. The band at 314 nm is mainly due to the transition
from HOMO-3 to LUMO, and the band at 261 nm is mainly
associated with the transition from HOMO-3 to LUMO+2 (see
Figure 2S, Supporting Information). It is known that the
presence of an electron-donating group has little effect on the
(π, π*) band near 260 nm but induces a strong absorption at
longer wavelengths.10 The 267 nm third harmonic of the
fundamental 800 nm light from the regenerative amplifier was
used as the pump laser pulse in the fs-TA experiments reported
here. The 266 nm fourth harmonic of the fundamental 1064
nm light from a Nd:YAG laser was employed for the pump
laser pulse in the ns-TA and ns-TR3 experiments reported here.
For all of the time-resolved experiments reported here, the
pump laser pulses excite the (π, π*) band near 260 nm.
The optimized structures of the TPA and TPA− ground
states are presented in Figure 5 (top) and (middle). Inspection
of Figure 5 reveals that there is a significant difference in the
C21−C27 bond length for TPA and TPA−, 1.526 and 1.627 Å,
respectively. The C21−C27 bond length for TPA− is longer than
that of TPA and appears to be close to that of a normal single
carbon−carbon bong length. This indicates the bond energy of
the C21−C27 bond for the TPA− form of the molecule may be
814
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
on different time scales, the spectra of the early (0.2−1, 1−11
ps) and later (after 11 ps) delay times are given separately. At
the beginning, a weak band assigned to intermediate 1 at about
340 nm grew in rapidly and red-shifted to 360 nm. This
indicates intermediate 1 decays to generate another inter-
mediate 2 within a very short time (see Figure 6a).
Subsequently, the absorption band at 360 nm begins to
decrease gradually and is accompanied by the generation of a
new absorption band at about 600 nm that is assigned to
intermediate 3 up to 11 ps. When the band at 600 nm reaches a
maximum, the band at 360 nm still does not completely decay
and this indicates that intermediate 3 also has some absorption
at about 360 nm. An isosbestic point was observed at around
420 nm between the bands at 360 and 600 nm, and this
suggests that intermediate 2 is the precursor for intermediate 3.
With the evolution of the time delay, intermediate 4 associated
with the absorption band around 415 nm was formed. The
growth of the absorption band at 415 nm supplies evidence for
the formation of intermediate 4, and two new isosbestic points
located at about 380 and 440 nm indicate that intermediate 4
was generated from the decay of intermediate 3.
Figure 4. Shown is a comparison of (a) the resonance Raman
spectrum of TPA obtained in a neutral PBS aqueous solution with (b)
the resonance Raman spectrum of TPA obtained in a basic aqueous
solution at pH 13.
The first change observed at very early times (before 1 ps) is
assigned to the internal conversion (IC) process from
intermediate 1 (assigned to the higher excited singlet Sn state
species) to intermediate 2 (assigned to the lowest lying exicted
singlet state S1). This assignment is consistent with the typical
IC photophysical process on a femtosecond time scale.41 Due
to the high efficiency of the intersystem crossing (ISC) for TPA
which can be up to 0.9,20 the subsequent process in the spectral
development in the fs-TA spectra can be reasonably attributed
to ISC from the S1 state to intermediate 3 that is assigned to
the lowest lying excited triplet state T1. As mentioned before,
carbon dioxide will leave spontaneously when we attempt to
optimize the triplet state TPA− species in some DFT
calculations, which is consistent with previously reported
DFT results of Eriksson and co-workers24 and also similar to
analogous computational results for KP.43 Therefore, we think
the triplet state T1 of TPA− (e.g., intermediate 3)
spontaneously decaboxylates to produce intermediate 4. DFT
calculations were performed to optimize the structure of
intermediate 4, and the results for its charge and spin
distribution are displayed in Figure 7. Examination of Figure
7 (left) reveals that much of the negative charge is located on
the O7 atom with some modest negative charge also located on
the C3, C1, and C8 atoms (see Figure 3S in the Supporting
Information for the labeling of the atoms). The positive charge
was mainly situated on the C6 atom with some positive charge
also located on the S2, C14, and C15 atoms. Further inspection
of Figure 7 (right) shows that around half of the spin density is
located on the C8 atom and most of the remaining spin is
located on the C3, C6, and O7 atoms (see Table 1S in the
Supporting Information for details). On the basis of these
results, it appears that intermediate 4 is a species that has
biradical character and is denoted as TB3 with a structure
shown in Figure 3S (Supporting Information).
Figure 5. Shown are the optimized geometry structures calculated at
the (U)B3LYP/6-311G(d,p) level of theory for the ground singlet
state of TPA (top), the ground state of TPA− (middle), and the triplet
state TPA− species (bottom).
significantly weaker and easier to cleave than for the TPA
neutral form of the molecule. When we attempted to optimize
the structure of the triplet state TPA− species, we found that
the cleavage of the bond located at C21−C27 took place
spontaneously (see Figure 5, bottom structure), whereas this
behavior was not observed for triplet state TPA species. These
results are consistent with those reported by Eriksson and co-
workers.24
fs-TA and ns-TA Spectroscopy Investigations on
TPA−. Figure 6 displays the fs-TA spectra obtained after 267
nm excitation of TPA in a neutral PBS aqueous solution from 0
to 3000 ps. To clearly indicate the spectral changes occurring
The assignment of TB3 to intermediate 4 was further
supported by a predicted absorption spectrum calculated at the
BPW91/CC-PVDZ level of theory. The predicted absorption
spectrum was depicted using Gausssum 2.244 with a fwhm of
7000 cm−1 and is shown in Figure 6d (see Table 3S, Supporting
Information, for details of these computational results).
Examination of Figure 6d shows that three strong absorption
bands are clearly seen and are located at 350, 450, and 680 nm,
815
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
Figure 6. (a−c) The fs-TA spectra obtained for TPA in a neutral PBS aqueous solution and (d) the predicted absorption spectrum of the TB3
intermediate from TD-DFT calculations done at the BPW91/CC-PVDZ level of theory (see text for details).
Figure 7. Mulliken charge distribution (left) and spin distribution (right) for the TB3 species calculated at the UB3LYP/6-311G(d,p) level of theory.
respectively. Taking into consideration that the current level of
theory may have an error of about 0.2 eV, the TD-DFT
predicted absorption spectrum can be regarded to be in
comparatively good agreement with the experimental absorp-
tion spectrum.
Table 1 provides a summary of the transient species observed
in the fs-TA spectra along with the tentative assignments of the
intermediates and their approximate absorption band maxima
observed in the experimental spectra.
To better understand the photophysical conversion processes
seen at early times in the fs-TA spectra, the kinetics were
monitored at 360 and 600 nm, respectively. The experimental
results and a fitting of the data at 360 nm within 20 ps are
shown in Figure 8a. Because both Sn and S1 have contributions
to the absorption band at 362 nm, a biexponential function was
employed to fit the kinetics a best fit found time constants of τ1
= 120 17 fs and τ2 = 2.7 0.06 ps. The very short growth
time (120 fs) is attributed to a fast IC process from Sn to S1.
Furthermore, as a result of the energies being close to each
other for the lowest lying excited singlet state and the excited
triplet, the ISC process is also easy to take place and we
attribute the time constant of 2.7 ps to the ISC process to
convert S1 to T1. To investigate the subsequent dynamical
change taking place after the ISC process produces T1, the
kinetics of the aborption band at 600 nm was monitored, and
its best-fit to the data is presented in Figure 8b. The growth and
decay of the spectral data at 600 nm were well fit by a
biexponential function with time constants of τ1 = 3.0 0.24 ps
and τ2 = 320 67 ps. The shorter time constant of 3.0 ps was
assigned to the formation time of T1 and agrees well with the
decay time constant of 2.7 ps for the S1 state monitored at 360
nm. The T1 state then appears to undergo a prompt
decarboxylation reaction to form TB3 with a time constant of
320 ps that was monitored at 600 nm. fs-TA experiments were
816
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
Figure 8. Kinetics and their fitting plots of the TPA fs-TA data in neutral PBS aqueous solution (a) at 360 nm within 20 ps and (b) at 600 nm within
3000 ps.
also done in pure MeCN for comparison purposes, and these
spectra are displayed in Figure 4S (Supporting Information)
along with the kinetics and relevant fitting plots of the
absorption band at 350 nm. MeCN is an inert solvent with a
lower polarity relative to pure water, so it can serve to stabilize a
(n, π*) triplet compared with a higher polarity water solvent.
The time constants observed after 267 nm excitation of TPA in
pure MeCN from a best-fit of a biexponential function to the
data were τ1 = 296 29 fs and τ2 = 2.4 0.17 ps, respecitvely.
The first time constant of τ1 = 296 fs can be attributed to IC
from Sn to S1, and the second time constant of τ2 = 2.4 ps can
be atttributed to ISC from S1 to T1. These two time constants
are practially identical with those obtained in a neutral PBS
aqueous solution which suggests there is not much solvent
polarity dependence for the IC from Sn to S1 and the ISC from
S1 to T1 for TPA. We note that many investigations on an
analogous NSAID drug, KP,42,45−47 indicate that photo-
decarboxylation prefers to proceed from the lowest lying triplet
state and the data above suggests that the TPA decarboxylation
reaction probably occurs in a similar manner.
that there is a (n, π*) transition character for T1 and a (π, π*)
transition character for T2. This is consistent with the
experimental results obtained from the fs-TA experiments
done in both pure MeCN solvent and in a neutral PBS aqueous
solution.
To investigate the photochemistry of TPA in neutral PBS
aqueous solution on the nanosecond and microsecond time
scales, ns-TA experiments were performed. Figure 10 presents
The first and second excited triplet state (T1 and T2)
geometries were optimized using the TD-DFT method, and the
frontier orbitals obtained are shown in Figure 9. Analysis of the
TD-DFT results indicates the HOMO and LUMO correspond
to the T1 transition and the T2 transition is associated with the
HOMO-1 to LUMO transition. Inspection of Figure 9 reveals
Figure 10. The ns-TA spectra obtained after 266 nm excitation of
TPA in a neutral PBS aqueous solution under nitrogen-saturated
conditions. Inset: An enlarged frame of spectra between 420 and 500
nm observed from 2 to 8 μs is also shown.
ns-TA spectra obtained after 266 nm excitation of TPA in a
neutral PBS aqueous solution under nitrogen-saturated
conditions. The profile of the spectrum at 0 ns delay time is
characterized by three absorption bands with maxima at about
360, 420, and 600 nm, and this absorption spectrum very
closely resembles the 3000 ps spectrum observed in the fs-TA
spectra. This implies the 0 ns spectrum in the ns-TA
experiments and the 3000 ps spectrum in the fs-TA
experiments are seeing the same transient species and have
the same assignment. With time evolution, the absorption band
at 360 nm increases along with a slight red shift to 369 nm, and
this is accompanied by a substantial decline of intensity in the
absorption band at 420 nm within 2 μs. Subsequently, the
absorption band at 369 nm begins to decrease gradually with a
marginal blue shift, and this is accompanied by a new
absorption band increasing in intensity at about 450 nm within
Figure 9. Shown are the DFT calculated frontier orbitals for the
excited state triplet TPA− involved in (a) the T1 HOMO, (b) the T1
LUMO; (c) the T2 HOMO-1, and (d) the T2 LUMO.
817
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
8 μs (see Figure 10, inset). Finally, both absorption bands at
about 360 and 450 nm decreased progressively and the decay
time for both occurred on the microsecond time scale. The
absorption bands observed at 0 ns delay time can be assigned to
the TB3 due to the close similarity with the fs-TA spectrum
observed at 3000 ps. The TB3 species can undergo protonation
to give a triplet neutral species that is denoted here as TBP3.
Since the pKa of around 8.222 for TB3 is higher than the pH
value of about 7.4 in the neutral PBS aqueous solution
examined here, the protonation of the TB3 species to form the
TBP3 species appears to be a likely process to observe. DFT
calculations were done for this protonation reaction and no
transition state (TS) was found, indicating the protonation
process proceeds spontaneously under relatively acidic
conditions. After the protonation reaction to make the TBP3
species, an ISC process can take place to produce the singlet
TBP species which has absorption bands at about 360 and 450
nm. The results of DFT calculations for the free energies of the
TBP3 and TBP species found that TBP has a lower energy than
TBP3 (see the Supporting Information for details). The
optimized structure of TBP is shown in Figure 11, and this
spectrum has a very strong absorption band at about 400 nm
and a weak band located at about 320 nm.
The TBP3 intermediate has an absorption band at about 369
nm and was generated from the TB3 intermediate by a
protonation reaction. Then, the TBP3 intermediate decays to
TBP by an ISC process and TBP has its main absorption band
at about 450 nm. Consequently, the growth time constant of
152 5 ns (see Figure 13a1) obtained from a monoexponential
function fitting at 360 nm within 600 ns can be attributed to the
formation time of the TBP3 species; the decay time constants of
17
0.3 nd 418
6 μs (see Figure 13b1) derived from a
biexponential function fitting at 360 nm from 2 μs to about 2.5
ms are assigned to the lifetimes of TBP3 and TBP
intermediates; the decay time constant of 457
8 μs (see
Figure 13c1) determined from a monoexponential function
fitting at 450 nm from 280 μs to about 2.5 ms is attributed to
the decay of the TBP species.
Investigation of the Transient Species and the Final
Product by ns-TR3 Spectroscopy. Photoinduced decarbox-
ylation has been found to be an efficient and general reaction
for different types of arylcarboxylic acids in aqueous solutions at
pH > pKa.12 In a prior section, a proposed photo-
decarboxylation mechanism that occurs in a neutral PBS
aqueous solution was given on the basis of the results of the fs-
TA and ns-TA experimental spectra. The transient absorption
spectra observed are relatively broad and featureless and similar
to one another, since many of the transient species have a
similar chromophore. ns-TR3 spectra can help provide new
insight into the identity, structure, and properties of several
transient species associated with the photo-decarboxylation
reaction observed in the degradation process of TPA. Figure 14
displays ns-TR3 spectra obtained using a 266 nm pump
wavelength and a 341 nm probe wavelength for an ∼1 mM
TPA neutral PBS aqueous solution.
Figure 11. The optimized geometry structure of TBP calculated at the
(U)B3LYP/6-311G(d,p) level of theory.
Inspection of Figure 14 shows that four species are generated
from 0 ns to 500 μs. In the early time spectra, only two main
Raman bands located at 1593 and 1325 cm−1 were observed. As
the delay time increases, the Raman band at 1325 cm−1
gradually decreases and then disappears, while the band at
1593 cm−1 increases in intensity and a new Raman band
appears at 1542 cm−1 band and increases in intensity within
about 1 μs. Subsequently, both Raman bands at 1542 and 1593
cm−1 begin to decrease in intensity accompanied by the
appearance of another new Raman band at 1635 cm−1
structure exhibits an obvious enol character for the TBP
intermediate. The predicted absorption spectrum of TBP3
obtained from the TD-DFT calculations performed at the
(U)B3LYP/6-311G(d,p) level of theory with consideration of
the solvation effect by means of the PCM model using UAHF
radii (Tables 4S and 5S, Supporting Information) is compared
with the experimental spectrum in Figure 12. Inspection of
Figure 12 reveals that the predicted TBP3 absorption spectrum
has one strong band located at 375 nm and one weak band
located at 552 nm. Similarly, the predicted TBP absorption
Figure 12. The predicted absorption spectra of the TBP3 (a) and TBP (b) species obtained from the TD-DFT calculations done at the (U)B3LYP/
6-311G(d,p) level of theory (see text for details).
818
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
Figure 13. The kinetics of absorption bands of TPA obtained in neutral PBS solution at 363 nm early stage, a1, fitting residual distribution, a2, at 363
nm late stage, b1, fitting residual distribution, b2, and at 450 nm, c1, fitting residual distribution, c2. Note: A is the fitting parameter.
generated and increase in intensity before 100 μs. At the
longest time delays examined, both Raman bands at 1542 and
1635 cm−1 decreased in intensity, while no obvious decrease in
intensity was observed for the band at 1593 cm−1. In addition, a
strong Raman band was observed at 1442 cm−1 that increased
in intensity from about 100 to 500 μs and this appears to
correspond to the decrease in intensity of the Raman bands at
1542 and 1635 cm−1 over the same time region. The changes
observed in the ns-TR3 spectra indicate that there are four
species present from 5 ns to 500 μs and the earlier species
appear to be the precursors for the later time species. The
analysis of the fs-TA and ns-TA results indicated an excited
triplet was involved in the photo-decarboxylation reaction, and
it is interesting to note this is different from the photo-
decarboxylation of o-acetylphenyl-acid that was found to start
from a highly excited singlet state with charge-transfer
character.48 This could possibly be due to the acid chain of
o-acetylphenyl-acid being much closer to the carboxyl group so
as to better facilitate an intramolecular proton transfer to the
carboxyl group in the (π, π*) state.
done to find the optimized geometry and vibrational
frequencies for likely species. The comparisons between some
of the experimental ns-TR3 spectra and the predicted Raman
spectra for some possible transient species like TB3 and TBP3
are shown in Figure 15. The Raman bands at 1541 and 1594
cm−1 are characteristic bands for the TBP3 species based on the
good agreement between the spectra shown in Figure 15a for
the predicted Raman spectrum of TBP3 and the 400 ns TR3
spectrum shown in Figure 15b. In addition, inspection of Figure
15b and c shows that the Raman bands at 1541 and 1594 cm−1
increase in intensity relative to the lower frequency Raman
bands in the 1100−1500 cm−1 region and also new Raman
bands appear in this spectral region with an increase in delay
time from 10 to 400 ns. This indicates that more than one
species coexist in the 10 and 400 ns spectra, and their relative
populations change substantially between 10 and 400 ns. The
400 ns TR3 spectrum appears to be mostly due to the TBP3
intermediate. The 10 ns TR3 Raman bands in the 1100−1500
cm−1 region shown in Figure 15c exhibit good agreement with
the vibrational frequency pattern for the predicted Raman
spectrum of the TB3 species shown in Figure 15d. This and the
noticeable presence of the 1541 and 1594 cm−1 Raman bands
To help identify what transient species can be produced from
the photo-decarboxylation reaction of TPA, calculations were
819
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
ns in Figure 15c and then the TB3 gradually undergoes
protonation to form TBP3 which becomes the predominant
species at 400 ns in the TR3 spectra shown in Figure 15b.
The calculated optimized structure of TBP3 is shown in
Figure 5S (Supporting Information), and the distribution of the
spin density is displayed in Figure 16. Inspection of Figure 16
Figure 16. The Mulliken spin distribution for the TBP3 species
calculated at the UB3LYP/6-311G(d,p) level of theory.
shows that the spin density is mainly located on the C12 and
C
21 atoms. Thus, it can be concluded that the TBP3 species has
a biradical character. A similar result has been suggested by
Eriksson and co-workers24 that a prompt photo-decarboxyla-
tion of the triplet conjugate base of TPA takes place to produce
a carbanion, and its biradical form is the more stable resonance
structure. Our TR3 results experimentally confirm that this is
the case.
Figure 14. The ns-TR3 spectra of 1 mM TPA in a neutral PBS
aqueous solution obtained using a 266 nm wavelength pump and a
341 nm wavelength probe.
Most of the Raman bands observed for the TBP3 species are
due to vibrations associated with the C−C stretch motions of
the phenyl rings. For example, the Raman bands located at 849,
1329, 1355, and 1575 cm−1 are mainly due to these kinds of
vibrational modes. The Raman band at 1171 cm−1 has
contributions not only from the C−C stretch motions but
also from a C−H bend motion. The calculated optimized
structure in Figure 5S (Supporting Information) for the TBP3
intermediate reveals a delocalization of the radical electrons
into the aromatic ring, making the C−C bond length change to
become about 1.4 Å, which stabilizes the triplet state. More
detailed information for the vibrational modes associated with
the observed Raman frequencies is presented in Table 8S
(Supporting Information) and compared to the corresponding
calculated vibrational frequencies for the TB3 and TBP3
intermediates.
Figure 15. Comparison of the DFT predicted spectrum for the TBP3
intermediate (a) and the TB3 intermediate (d) with the experimental
ns-TR3 spectra of TPA in a neutral PBS aqueous solution acquired at a
time delay of 10 ns (c) and 400 ns (b).
On the basis of the conclusions of ns-TA spectra that TBP3
will decay to its singlet form TBP, the predicted normal Raman
spectrum of TBP was compared with the experimental
spectrum obtained at a 120 μs delay time and the comparison
of these spectra is displayed in Figure 17. Examination of Figure
17 reveals the good agreement between the vibrational
frequency patterns in the two spectra. The differences in the
relative intensities between the experimental and calculated
spectra are attributed to the TR3 spectrum being resonantly
enhanced, whereas the DFT calculations are for the normal
Raman spectrum that is not resonantly enhanced. More
detailed information for the comparison and assignment of
experimental and calculated Raman bands are given in Table 9S
of the Supporting Information.
Most of the Raman bands observed for the TBP species are
due to the vibrations associated with the C−C stretch motions
of the phenyl rings. For example, the Raman bands at 1329,
1609, and 1635 cm−1 are mainly due to these kinds of
vibrational modes. The Raman band at 1635 cm−1 has
contributions not only from the C−C stretch modes but also
that are characteristic bands for the TBP3 species indicate that
both the TB3 and TBP3 intermediates noticeably contribute to
the experimental 10 ns TR3 spectrum. The ns-TA spectra along
with the predicted absorption spectra also indicated that the
TB3 undergoes a protonation process to produce the TBP3
intermediate on a similar time scale. The TR3 spectral results
here are consistent with the ns-TA results. The ns-TR3 results
also provide (to our knowledge) the first time-resolved
vibrational spectroscopic characterization of these two key
intermediates and confirm that the protonation reaction takes
place on the nanosecond time scale in TPA after the
decarboxylation on the picosecond time scale produces the
TB3 biradical intermediate. In summary, during the early
nanosecond delay times, the TB3 and TBP3 intermediates
coexist and the former one was the predominant species at 10
820
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
DFT calculations were done to estimate the activation energy
for the reaction of TBP to form the DTPA final product in the
gas phase and in aqueous solution. These DFT calculations
found that the reaction barriers were 19.8 kcal/mol in the gas
phase and 17.9 kcal/mol (see Figure 19) with consideration of
Figure 17. Comparison of the experimental ns-TR3 spectrum of TPA
in a neutral PBS aqueous solution acquired at a time delay of 120 μs
(top) with the DFT predicted spectrum for the TBP intermediate
(bottom).
Figure 19. The reactive energy profile obtained from the DFT
calculations for isomerization in gas and aqueous phases.
from an O−H bend motion. The results for the ns-TA and the
TR3 spectra both indicate that, within a 100 μs delay time, the
TB3 intermediate undergoes a fast protonation process to
generate the TBP3 intermediate that also has some biradical
character like the TB3 intermediate. The TBP3 intermediate
subsequently undergoes ISC to form the singlet TBP
intermediate.
In view of the instability of TBP that has appreciable enol
character, one could expect further reaction may take place to
produce the DTPA final product. Further inspection of Figure
14 shows that an obvious new Raman band is generated at 1442
cm−1 and this can be attributed to a final product DTPA
characteristic Raman band. Figure 18 presents a comparison
a solvation effect by means of the PCM using UFF radii in
aqueous solution (see the Supporting Information for details of
the calculations). The optimized geometries of the reactant
complex (RC), transient state (TS), and product complex (PC)
of the reaction pathway are shown in Figure 6S (Supporting
Information), and the structures are labeled by selected bond
length values (in Å). The reaction barriers are reasonably
consistent with the kinetics time constant in the 400−500 μs
range for the formation of the DTPA final product in the ns-TA
and ns-TR3 spectra. We also performed DFT calculations to
estimate the absorption spectrum of DTPA using the TD-
B3LYP/6-311G(d,p) level of theory and consideration of the
solvation effect by means of PCM using UAHF radii (see
Figure 7S and Table 6S, Supporting Information, for the
calculation results), and we found that DTPA has a larger
oscillator strength estimated absorption band at 341 nm which
is consistent with the ns-TA and ns-TR3 spectra results.
Scheme 3 presents a reaction mechanism for the photo-
chemistry of TPA− in a near neutral PBS aqueous solution
based on our present time-resolved spectroscopy experiments
that directly observed and characterized key reaction
intermediates from femtoseconds to the formation of the
DTPA final product. Scheme 4 presents a similar reaction
mechanism based on time-resolved spectroscopic observations
for the photochemistry of KP− in a near neutral PBS aqueous
solution. Comparison of Schemes 3 and 4 shows that, while the
sequence of events for the TPA− and KP− drugs in a near
neutral PBS aqueous solution are similar to one another, there
are some significant differences when the phenyl ring
connecting the carbonyl group to the carboxyl containing
side chain in KP− is replaced by a thiophene ring in TPA−. For
example, the ISC time becomes faster from S1 to T1 to have a
rate constant of 3.0 ps for TPA− compared to a rate constant of
about 8.8 ps for KP−. This can be partially attributed to a heavy
atom effect due to the presence of sulfur (S) in the thiophene
ring of TPA−. An intriguing effect is that the decarboxylation
time constant is also faster in TPA− (about 320 ps) compared
to KP− (about 568 ps), and this may be related to the
thiophene stabilization of the TB3 biradical intermediate
(lifetime of about 152 ns) produced from the decarboxylation
of the triplet TPA− compared to the 3BC− biradical
Figure 18. Shown is a comparison of the experimental ns-TR3
spectrum of TPA in neutral PBS solution acquired at a time delay
of 500 μs (top) with the DFT calculated normal Raman spectrum for
the DTPA final product (bottom).
between the experimental ns-TR3 spectrum at a 500 μs delay
time and a DFT calculated Raman spectrum for the DTPA final
product. The vibrational frequency patterns between the
experimental 500 μs ns-TR3 spectrum and the DFT calculated
spectrum for the DTPA in Figure 18 are in good agreement
with one another and indicate the 500 μs ns-TR3 spectrum is
mainly due to the DTPA final product. More detailed
information for the experimental and computational Raman
DTPA spectral results is given in Table 10S of the Supporting
Information.
821
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
Scheme 3. Proposed Reaction Mechanism for the Ultraviolet Excitation of TPA in a Near Neutral PBS Aqueous Solution Based
on the Direct Time-Resolved Spectroscopy Observations of This Study from Femtoseconds to the Formation of the DTPA
Final Product
Scheme 4. Proposed Reaction Mechanism of KP after Ultraviolet Photo-Excitation in a Near Neutral PBS Aqueous Solution
Based on the Direct Time-Resolved Spectroscopy Observations in the Literature27,31
intermediate (lifetime of about 18 ns) formed from the
decarboxylation of the triplet KP−. This stabilization of the
thiophene containing biradical anion intermediate from the
decarboxylation reaction appears linked to a slower protonation
of this biradical which results in a longer lifetime. The
protonated biradical species containing a thiophene moiety
(TBP3) also has a very long lifetime of about 17 μs compared
to the lifetime of about 411 ns for the analogous protonated
biradical (3BCH) associated with the photochemistry of KP−
that has a phenyl ring instead of thiophene. The thiophene
moiety leads to substantially longer lifetimes for both the anion
biradical and its protonated biradical species. It will be
interesting in the future to study the chemical reactivity of
these biradical species that contain a thiophene versus a phenyl
ring connecting the carbonyl group to the carboxyl containing
side chain in aromatic carbonyl NSAIDs. In so far as the
biradical intermediates are linked to phototoxicity of these
NSAIDs, their properties and chemical reactivity to biological
components should provide additional insights into the
phototoxicity of these kinds of drugs.
spectroscopy methods and also using DFT calculations. In
the fs-TA spectra, a strong transient absorption band at 358 nm
associated with the TPA− Sn → S1 was observed and then the
S1 intermediate underwent a fast ISC process to mainly
transform to the T1 state, which is characterized by two
transient absorption bands at 360 and 600 nm. Subsequently,
T1 underwent decarboxylation to generate a triplet biradical
species TB3, which was clearly detected in the fs-TA and ns-TA
spectra. Due to the higher pKa value of TB3 than the pH value
of the near neutral PBS aqueous solution, protonation appeared
to take place easily to produce the neutral species TBP3. In
both the ns-TA and ns-TR3 spectra, the growth of TBP3 was
directly observed and then the TBP3 intermediate decayed
mostly by ISC to produce its singlet form TBP. The ns-TA and
ns-TR3 spectra also directly observed the reaction of the TBP
intermediate to form the final product DTPA. The time-
resolved spectroscopy experiments presented here have
observed all of the intermediate species from femtoseconds
to production of the major final product DTPA for the
photochemistry of TPA− in a near neutral PBS aqueous
solution. The ns-TR3 experiments also have provided the first
time-resolved vibrational spectroscopic characterization of
several key intermediates to clearly identify them and gain
more insight into their structure and properties. Comparison of
the time-resolved spectroscopy results here for TPA− to those
reported in the literature for the related KP− compound under
CONCLUSION
■
The photo-decarboxylation and overall reaction mechanism for
the photochemistry of TPA in a near neutral PBS aqueous
solution was explored using fs-TA, ns-TA, and ns-TR3
822
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
similar conditions indicate that the thiophene moiety in TPA−
that replaces one phenyl ring in KP− leads to a faster ISC from
S1 to T1 and also a faster decarboxylation of the T1 intermediate
to produce the anionic biradical intermediate (TB3 in Scheme 3
(6) Agapakis-Causse, C.; Bosca, F.; Castell, J. V.; Hernandez, D.;
Marin, M. L.; Marrot, L.; Miranda, M. A. Photochem. Photobiol. 2000,
71, 499−505.
(7) Bosca, F.; Miranda, M. A.; Morera, I. M.; Samadi, A. J. Photochem.
Photobiol., B 2000, 58, 1−5.
3
compared to BC− in Scheme 4). In addition, the lifetimes of
(8) Miranda, M. A.; Lahoz, A.; Bosca, F.; Metni, M. R.; Abdelouahab,
F. B.; Castell, J. V.; Perez-Prieto, J. Chem. Commun. 2000, 22, 2257−
2258.
the anionic biradical intermediate and its protonated biradical
intermediate is much longer for the TPA− associated species
(TB3 and TBP3 in Scheme 3) than the KP− ones (3BC− and
3BCH in Scheme 4).
(9) Romani, A.; Ortica, F.; Favaro, G. J. Photochem. Photobiol., A
2000, 135, 127−134.
(10) Sortino, S.; Cosa, G.; Scaiano, J. C. New J. Chem. 2000, 24, 159−
163.
(11) Starrs, S. M.; Davies, R. J. H. Photochem. Photobiol. Sci. 2000, 72,
291−297.
ASSOCIATED CONTENT
* Supporting Information
■
S
The TD-DFT predicted absorption spectrum of the TPA−
species is given. The fs-TA spectra of TPA in MeCN solvent
are provided. The DFT calculated frontier orbitals of TPA−
involved in the HOMO-3, LUMO, and LUMO+2 orbitals are
given. The chemical structure of TB3 is given. The fs-TA
spectra and their kinetics monitored at 352 nm of TPA in a
pure MeCN solvent are shown. The DFT optimized structure
of TBP3 is given. The optimized geometries of the RC, TS, and
PC obtained from the DFT calculations for the reaction of TPA
are given. The DFT predicted absorption spectrum of the
DTPA final product is given. The DFT Mulliken charge
distribution and spin distribution of the TB3 species are given.
Excited-state energies and oscillator strengths determined from
DFT calculations are shown for the ground state TPA−, the
TB3 intermediate, the TBP3 intermediate, the TBP species, and
the final product are given. Comparisons of the experimental
resonance Raman or ns-TR3 spectra vibrational frequencies
with the DFT calculated vibrational frequencies for the TPA−
species, the TB3 and TBP3 intermediates, the TBP species, and
the DTPA final product with preliminary vibrational assign-
ments and qualitative descriptions of the vibrational modes in
the 800−1800 cm−1 region are presented. The Cartesian
coordinates used in the calculations of all the intermediates are
provided. The full ref 39 is given. This material is available free
of charge via the Internet at http://pubs.acs.org.
(12) Xu, M. S.; Wan, P. Chem. Commun. 2000, 21, 2147−2148.
(13) Bosca, F.; Marin, M. L.; Miranda, M. A. Photochem. Photobiol.
Sci. 2001, 74, 637−655.
(14) Miranda, M. A. Pure Appl. Chem. 2001, 73, 481−486.
(15) Monti, S.; Encinas, S.; Lahoz, A.; Marconi, G.; Sortino, S.;
Perez-Prieto, J.; Miranda, M. A. Helv. Chim. Acta 2001, 84, 2452−
2466.
(16) Samadi, A.; Martinez, L. A.; Miranda, M. A.; Morera, I. M.
Photochem. Photobiol. Sci. 2001, 73, 359−365.
(17) Sortino, S.; Martinez, L. J.; Marconi, G. New J. Chem. 2001, 25,
975−980.
(18) delaPena, D.; Marti, C.; Nonell, S.; Martinez, L. A.; Miranda, M.
A. Photochem. Photobiol. Sci. 1997, 65, 828−832.
(19) Martinez, L. J.; Scaiano, J. C. J. Am. Chem. Soc. 1997, 119,
11066−11070.
(20) Encinas, S.; Miranda, M. A.; Marconi, G.; Monti, S. Photochem.
Photobiol. Sci. 1998, 67, 420−425.
(21) Bosca, F.; Miranda, M. A. Photochem. Photobiol. Sci. 1999, 70,
853−857.
(22) Encinas, S.; Miranda, M. A.; Marconi, G.; Monti, S. Photochem.
Photobiol. Sci. 1998, 68, 633−639.
(23) Vinette, A. L.; McNamee, J. P.; Bellier, P. V.; McLean, J. R. N.;
Scaiano, J. C. Photochem. Photobiol. Sci. 2003, 77, 390−396.
(24) Musa, K. A. K.; Eriksson, L. A. J. Phys. Chem. B 2009, 113,
11306−11313.
(25) Ma, J. N.; Cheng, S. C.; An, H. Y.; Li, M. D.; Ma, C. S.; Rea, A.
C.; Zhu, Y.; Nganga, J. L.; Dore, T. M.; Phillips, D. L. J. Phys. Chem. A
2011, 115, 11632−11640.
AUTHOR INFORMATION
Corresponding Author
*Fax: (+852) 2597-1586. Phone: (+852) 2859 2160. E-mail:
phillips@hku.hk.
■
(26) Ma, J. N.; Li, M. D.; Phillips, D. L.; Wan, P. J. Org. Chem. 2011,
76, 3710−3719.
(27) Li, M. D.; Ma, J. N.; Su, T.; Liu, M. Y.; Yu, L. H.; Phillips, D. L.
J. Phys. Chem. B 2012, 116, 5882−5887.
Notes
(28) Ma, J. N.; Rea, A. C.; An, H. Y.; Ma, C. S.; Guan, X. G.; Li, M.
D.; Su, T.; Yeung, C. S.; Harris, K. T.; Zhu, Y.; Nganga, J. L.; Fedoryak,
O. D.; Dore, T. M.; Phillips, D. L. Chem.Eur. J. 2012, 18, 6854−
6865.
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
(29) Ma, C. S.; Kwok, W. M.; Chan, W. S.; Du, Y.; Zuo, P.; Kan, J. T.
W.; Toy, P. H.; Phillips, D. L. Curr. Sci. 2009, 97, 202−209.
(30) Li, M. D.; Du, Y.; Chuang, Y. P.; Xue, J. D.; Phillips, D. L. Phys.
Chem. Chem. Phys. 2010, 12, 4800−4808.
This work was supported by a grant from the Research Grants
Council of Hong Kong (HKU 7035/08P) and the University
Grants Committee Special Equipment Grant (SEG-HKU-07)
to D.L.P. Support from the University Grants Committee Areas
of Excellence Scheme (AoE/P-03/08) is also gratefully
acknowledged.
(31) Li, M. D.; Yeung, C. S.; Guan, X.; Ma, J.; Li, W.; Ma, C. S.;
Phillips, D. L. Chem.Eur. J. 2011, 17, 10935−10950.
(32) Ma, C. S.; Kwok, W. M.; An, H. Y.; Guan, X. G.; Fu, M. Y.; Toy,
P. H.; Phillips, D. L. Chem.Eur. J. 2010, 16, 5102−5118.
(33) Srinivasan, A.; Kebede, N.; Saavedra, J. E.; Nikolaitchik, A. V.;
Brady, D. A.; Yourd, E.; Davies, K. M.; Keefer, L. K.; Toscano, J. P. J.
Am. Chem. Soc. 2001, 123, 5465−5472.
REFERENCES
■
(1) Ljunggren, B. Photodermatology 1985, 2, 3−9.
(2) Przybilla, B.; Schwab-Przybilla, U.; Ruzicka, T.; Ring, J.
Photodermatology 1987, 4, 73.
(34) Chan, P. Y.; Ong, S. Y.; Zhu, P. Z.; Zhao, C. Y.; Phillips, D. L. J.
Phys. Chem. A 2003, 107, 8067−8074.
(3) Castell, J. V.; Gomezlechon, M. J.; Grassa, C.; Martinez, L. A.;
Miranda, M. A.; Tarrega, P. Photochem. Photobiol. 1994, 59, 35−39.
(4) Miranda, M. A.; Castell, J. V.; Gomezlechon, M. J.; Hernandez,
D.; Martinez, L. A. Toxicol. In Vitro 1995, 9, 499−503.
(5) Castelli, F.; De Guidi, G.; Giuffrida, S.; Miano, P.; Sortino, S. Int.
J. Pharm. 1999, 184, 21−33.
(35) Cances, E.; Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 107,
3032−3041.
(36) Mennucci, B.; Tomasi, J. J. Chem. Phys. 1997, 106, 5151−5158.
(37) Cossi, M.; Barone, V.; Mennucci, B.; Tomasi, J. Chem. Phys. Lett.
1998, 286, 253−260.
823
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824

The Journal of Physical Chemistry B
Article
(38) Tomasi, J.; Mennucci, B.; Cammi, R. Chem. Rev. 2005, 105,
2999−3093.
(39) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; et al. Gaussian 03; Gaussian, Inc.:
Wallingford, CT, 2004.
(40) Fillet, M.; Bechet, I.; Piette, V.; Crommen, J. Electrophoresis
1999, 20, 1907−1915.
(41) Modern Molecular Photochemistry; Turro, N. J., Ed.; University
Science Books: Sausalito, CA, 1991.
(42) Chuang, Y. P.; Xue, J. D.; Du, Y.; Li, M. D.; An, H. Y.; Phillips,
D. L. J. Phys. Chem. B 2009, 113, 10530−10539.
(43) Musa, K. A. K.; Matxain, J. M.; Eriksson, L. A. J. Med. Chem.
2007, 50, 1735−1743.
(44) O’Boyle, N. M.; Tenderholt, A. L.; Langner, K. M. J. Comput.
Chem. 2008, 29, 839−845.
(45) Suzuki, H.; Suzuki, T.; Ichimura, T.; Ikesue, K.; Sakai, M. J. Phys.
Chem. B 2007, 111, 3062−3068.
(46) Suzuki, T.; Okita, T.; Osanai, Y.; Ichimura, T. J. Phys. Chem. B
2008, 112, 15212−15216.
(47) Suzuki, T.; Osanai, Y.; Isozaki, T. Photochem. Photobiol. 2012,
88, 884−888.
(48) Ding, L.; Chen, X. B.; Fang, W. H. Org. Lett. 2009, 11, 1495−
1498.
824
dx.doi.org/10.1021/jp310315f | J. Phys. Chem. B 2013, 117, 811−824