View Article Online / Journal Homepage / Table of Contents for this issue

PCCP

Dynamic Article Links

Cite this: Phys. Chem. Chem. Phys., 2012, 14, 89038909

PAPER

www.rsc.org/pccp

On the uorescence of methyl salicylate: the signicance of its IMHB

Published on 11 January 2012. Downloaded by FAC DE QUIMICA on 24/04/2015 18:19:52.

J. Catalan*
Received 25th November 2011, Accepted 10th January 2012
DOI: 10.1039/c2cp23742c
The two forms of methyl salicylate bearing an intramolecular hydrogen bond (IMHB) are
responsible for the three uorescence emissions produced by this compound on electronic
excitation in inert media. Whereas the form possessing an IMHB between its hydroxyl group and
ether oxygen undergoes no excited state intramolecular proton transfer (ESIPT) in its rst excited
electronic state, that with an IMHB involving the carbonyl oxygen exhibits ESIPT with
near-unity eciency. Whereas the former species exhibits standard photophysical behaviour, the
latter species exhibits two uorescence emissions from the same electronic excited state; a
photophysical scheme is proposed, which brings together all the available photophysical evidence
for methyl salicylate in inert media.

Introduction
1

In 1956, Weller used his knowledge that electronic excitation

of an aromatic system increased the acidity of a hydroxyl
substituent by 68 pK units and the basicity of a carbonyl
group by about one unit,24 to study the photophysical
implications of the presence of a hydroxyl substituent and a
carbonyl substituent at neighbouring positions, and hence
forming an intramolecular hydrogen bond (IMHB), using
methyl salicylate (MS; Scheme 1) as a model compound. In
Fig. 2 of his seminal work, Weller illustrated the uorescent
behaviour in methylcyclohexane at 25 1C of methyl-2-methoxybenzoate (I in Scheme 1), which forms no IMHB, and methyl-2hydroxybenzoate (MS), which forms a strong IMHB; whereas
the former compound exhibits a single uorescence emission at

Scheme 1
Departamento de Qumica Fsica Aplicada, Universidad Autonoma de
Madrid, Cantoblanco, 28049 Madrid, Spain.
E-mail: Javier.catalan@uam.es

This journal is

the Owner Societies 2012

352 nm, the latter have two: one similar to that of the previous
compound at 361 nm and a stronger, anomalously red-shifted one
at 442 nm. Weller used the uorescence at 442 nm as evidence of a
proton transfer from the hydroxyl group to the carbonyl group in
MS in its rst excited electronic state. The two emissions were
quenched by both oxygen and CS2, so Weller concluded that the
two forms of MS, in equilibrium, were responsible for its uorescence emissions and proposed his well-known photophysical
scheme for this compound (see Fig. 3 of his paper1). The evidence
reported by Weller1,5,6 has provided the foundation for a major
area of research: excited state intramolecular proton transfer
(ESIPT),7 which has been used to develop interesting substances
such as laser dyes,8,9 UV lters,10,11 photostabilizers,12,13 photoactive proteins14,15 and biomolecular optical probes.16
According to Weller, the potential energy surface (PES) for
the rst excited electronic state of MS should contain two
minima, namely: one for a form similar to the FC structure
generated from a MS geometry which we shall designate the
normal form (MSn* in Scheme 1) and the other generated
from the normal form by transfer of the hydroxyl proton to
the carbonyl group which we shall call the transferred form
(MSt* in Scheme 1). According to Weller, the two forms are in
equilibrium in the excited electronic state.
Such a suggestive dual-uorescence photophysical scheme,
which produces a two-minimum PES, soon attracted interest
in its theoretical study and was assessed in the 1970s by using
the semi-empirical method CNDO/2,17,18 which revealed that,
in fact, the PES contains two minima but the energy dierence
between MSn* in its excited electronic state and the barrier to
the proton transfer is only 2.9 kcal mol 1 (i.e. possibly similar
to or even lower than the zero-point energy for the structure).
Therefore, this might be a barrierless process, similarly as
found by Weller et al.,6 which calculated an activation energy
of r0.12 kcal mol 1 for the process. The practical interest of
this nding led Goodman and Brus,19 of Bell Laboratories, to
Phys. Chem. Chem. Phys., 2012, 14, 89038909

8903

Published on 11 January 2012. Downloaded by FAC DE QUIMICA on 24/04/2015 18:19:52.

View Article Online

study MS isolated in a solid neon host at 4.2 K and to

conclude that there was no evidence of a dual minimum in
the rst excited singlet of the compound (i.e. that the process
was barrierless). Further evidence for a barrierless process was
subsequently provided by theoretical calculations.20
The interesting dual uorescence of MS exposed by Weller1
was examined in 1974 by Klopfer and Naundorf.21 They
found such dual uorescence to occur in methanol, but not
in methylcyclohexane (Mecyh), which led them to ascribe this
result to an equilibrium situation in the ground electronic state
of the compound. In 1976, Sandros22 showed MS in cyclohexane,
ethanol, methanol or 2,2,2-triuoromethanol to emit dierently
upon excitation at 286 and 316 nm, and also that the relative
intensity of the emission at 340 and 450 nm was correlated with
the well-known stability of the complexes formed by these
solvents and an ester; accordingly, the dual uorescence of MS
was ascribed to the presence of two dierent structures in the
ground state, namely: the normal structure (MSn, which was
designated the cis form), and the open structure containing no
hydrogen bond owing to a torsion of 1801 in its hydroxyl group,
which was designated the trans form (see Scheme 1). In 1978,
Smith and Kaufmann23 found (a) that the intensity of the
uorescence emission at 340 nm remained constant from 333 to
273 K in methylcyclohexane, whereas that at 450 nm increased
markedly between the two temperatures in the same solvent;
(b) that the excitation spectra were identically shaped but that
corresponding to lem = 340 nm (lmax = 330 nm) was blue-shifted
with respect to that at lem = 450 nm (lmax = 450 nm); and
(c) that MS had a single lifetime (280 ps) in methylcyclohexane
at 293 K but two in acetonitrile (1 ns at 340 nm and 0.1 ns at
450 nm). Two years later, Ford et al.24 showed (a) that the UV
emissions of MS at 340 nm in cyclohexane was not keeping
its constant intensity as previously assumed by Smith and
Kaufmann23, but rather increased signicantly with increasing
temperature, which they ascribed to a shift between the two
IMHB forms of MS (MSn and MSr in Scheme 1) in equilibrium
in the ground electronic state; (b) that, based on the fact that
carbon tetrachloride quenches the UV emission of MS dissolved
in methanol and cyclohexane, MS occurred not only in the
normal form, MSn, but also in at least two additional forms: an
open, trans, form and another (MSr) with the IMHB between
the hydroxyl group and the oxygen atom in the ester group.
Simultaneously, Acuna et al.25 concluded that the UV emission
of MS in hydrocarbon solvents was due to the presence of the
MSr form. Subsequently, Toribio et al.26 and Heimbrook
et al.27 conrmed the presence of MSr by IR and free-jet
spectroscopy, respectively.
As regards the presence of MS isomers in the ground state,
in 2007 Melandri et al.28 used free-jet rotational spectra for
MS to conclude that MSn exists as a single conformer and
calculated the other conformer (MSr) to be 2.4 kcal mol 1
more unstable under these conditions. Two years later, Massaro
et al.29 studied six potential MS isomers at the DFT/B3PW91 level
with exclusion of local and non-local correlation functionals, but
using a triple-z basis set with s,p,d polarization functions and extra
diuse d functions (6-311++G). They concluded that MSn was
the most stable conformer and that MSr was only 1.9 kcal mol 1
less stable, but emphasized the fact that conversion between
the two was subject to a high energy barrier (ca. 15 kcal mol 1)
8904

Phys. Chem. Chem. Phys., 2012, 14, 89038909

and hence highly unlikely. However, Pimentel et al.30 trapped

MS in Ar, Xe and SF6 matrices at 12 K and found MSn to be
photochemically converted into MSr and the process to be
highly irreversible. However ecient, the process requires the
participation of the rst triplet state of the compound.
Coe et al.31 used ab initio excited-state molecular dynamics
calculations with a multireference perturbation theory of the
electronic structure in the ESIPT process in MS and proposed
the presence of an S1/S0 minimal energy conical intersection
accounting for the experimental observation of a sharp
decrease in uorescence quantum yield at excitation energies
more than 1300 cm 1 above the excited-state origin.3234
Aquino et al.35 assessed the proton transfer curve for the
rst excited p,p* state of salicylic acid at the TDDFT and RI-CC2
levels, and concluded that the proton-transfer curves for the p,p*
state are very at. The assignment of an energy minimum structure
is quite arbitrary. Very recently, Massaro and Blaisten-Barojas36
examined the photophysics of the MS isomers MSn, MSt and
MSr by using time-dependent density functional theory in
combination with a large basis set and concluded that the dual
uorescence of MS was, as initially suggested by Weller,1 due to
the MSn* form (UV emission) and the MSt* form (blue emission).
Also, the temperature-dependence of this dual uorescence was
consistent with the results of Law and Shoham37 in methanol, who
made a very interesting proposal in this respect: that the barrier to
proton transfer in MS is temperature-dependent and must therefore inuence both uorescence emissions.
Based on the foregoing, one can be tempted to approach the
situation as follows: the two uorescence emissions of MS
observed in inert solvents can be assigned to two dierent
forms (MSn and MSr) in equilibrium in the ground electronic
state, both having an IMHB and hence very similar energies;
or, as stated by other authors, the emissions are associated to a
two-minimum PES in the 1(p,p*) state due to MSn* and
MSt*, respectively. Based on the former hypothesis, slowly
lowering the temperature of an MS solution will displace the
equilibrium in the ground electronic state to the more stable
form (MSn) and cause the corresponding UV emission of
MSr to gradually disappear. Based on the latter hypothesis,
lowering the temperature and viscosity of the medium will
generate a potential barrier that will gradually decrease the
uorescence of MSt* and increase that of MSn* as in the dual
proton transfer of the 7-azaindole dimer.40 Interestingly, the
latter possibility has never to date been systematically explored
and no study of the signicance of the solvent viscosity to the
equilibrium of the two forms, MSn and MSr, has been
reported. Rather, there have been disparate results such as
those of Smith and Kaufmann,23 who found the intensity of
the emission at 340 nm in methylcyclohexane to remain
constant from 333 to 273 K; those of Ford et al.,24 who found
such emission to increase with increasing temperature; or the
above-described reports that the two forms can hardly interconvert.28,29,32 Also, there have been systematic studies on the
inuence of the dipolarity of the medium on the ESIPT mechanism
in MS when its transferred form, MSt , has traditionally been
deemed zwitterionic.
In this work, we studied the uorescence emissions of MS in
three dierent inert media [2-methylbutane (2MB), 1-chlorobutane
(ClB) and squalane (SQ)] to elucidate the eect of temperature
This journal is

the Owner Societies 2012

Published on 11 January 2012. Downloaded by FAC DE QUIMICA on 24/04/2015 18:19:52.

View Article Online

on the equilibrium between the MSn and MSt forms, the

inuence of the dipolarity of the medium38,39 on the ESIPT
processwhich, as noted earlier, involves the conversion of a
normal form into a seemingly zwitterionic formand the
potential eect of the viscosity medium on ESIPT in the
excited electronic state and equilibrium between the forms
MSn and MSrwhich are mutually convertible by torsion
about the carbonyl group in the compound, respectively. In
addition, the solvents used allowed the generation of a
potential barrier to the ESIPT mechanism like that previously
found for the 7-azaindole dimer40 to be investigated.

Experimental section
2-Methylbutane (2MB) in Uvasol grade (moisture content r
0.005%) was obtained from Merck. 1-Chlorobutane (ClB) in
Chromasolv grade (moisture content r 0.001%) and 99.8%
purity, and squalane (SQ) of 99% purity were both supplied
by Aldrich. Squalane was passed through an SiO2 column
prior to use. MS solutions in 2MB and ClB were prepared in
such a way as not to exceed an optical density of 0.1 at the
excitation wavelength (i.e. at concentrations well below 10 4 M).
Sample temperatures ranged from 77 to 293 K and were
adjusted via an Oxford DN11704 cryostat equipped with an
ITC4 controller for interfacing to the spectrophotometers. The
cryostat was purged with dried nitrogen of 99.99% pure.
Sample temperatures from 343 to 293 K were maintained
within  0.1 1C with a Fisons Haake D8 GH thermostat.
All absorption, emission and excitation spectra were recorded
at a variable temperature, using Suprasil cells of 1 cm light
path xed to the cryostat. UVVis spectra were recorded on a
Cary-5 spectrophotometer.
Corrected excitation spectra were directly recorded on an
AB2 spectrouorimeter. A small fraction of the light intensity
used for excitation was switched to a Hamamatsu S1336-8BQ
photodiode by means of a beam splitter. Plotting the photodiode
sensitivity as a function of wavelength allowed us to characterize
changes in incident light intensity at each wavelength. The ratio
of emission intensity at the monitored wavelength and the
corresponding excitation wavelength was used to construct
absolute excitation spectra.
Fluorescence decays were determined by tting the AB2
uorimeter to the cryostat. The exciting light was supplied by
a LED and impinged on the side opposite that of excitation in
the uorimeter. Emission decays were determined on the side
opposite that of emission. The experimental set-up was controlled
with an EasyLife II electronic system from PTI, which aords
ecient processing of acquired data. The exciting LED used
delivered light of 310 nm. Quantum yields were determined by
using quinine bisulfate in 0.1 N H2SO4 (FF = 0.5141) as a
reference substrate.

same range, and that of the viscosity in SQ at 293193 K.

Finally, we shall examine the ESIPT process in the previous,
highly dierent media and propose a model consistent with the
overall experimental photophysical evidence for this compound.
UVVis spectroscopy of MS in 2MB
Fig. 1 shows the absorption spectra for an MS solution in
2MB at 293 to 113 K (i.e. in the liquid phase). As usual, the
spectral intensity signicantly increases in intensity as the
solution temperature is lowered and no signicant spectral
shift is observed. Interestingly, these spectra include an isosbestic
point at around 330 nm above and below which the intensity
decreases and increases, respectively, as the temperature is
lowered; also, the envelopes exhibit no vibronic structure at
any temperature. Initially, we have no clear explanation for
this spectral behaviour.
Fig. 2 shows the emission spectra for an MS solution in
2MB at 293113 K as obtained at lex = 290 nm. The spectra
exhibit a maximum at around 450 nm; interestingly, their
onsets overlap with the corresponding absorption onsets. As
can be seen, the emissions increase markedly in intensity as the
temperature is lowered; also, below 193 K, the samples exhibit
incipient vibronic structure at wavelengths lower than 450 nm
that is inappreciable at higher temperatures. It should be noted
that the gain in vibronic structure reected in the emission
spectra recorded at temperatures below 193 K is absent from
the absorption spectra of Fig. 1 and the corresponding excitation spectra. Fig. 3 shows selected normalized excitation
spectra obtained by monitoring light across the emission band

Fig. 1 UV absorption bands of MS dissolved in 2MB measured at

temperatures over the ranges 293113 K.

Results and discussion

Let us begin with the spectral behaviour of MS in such a highly
inert solvent as 2MB and examine the eect of temperatures
between 293 and 113 K on the photophysics of this compound,
and then examine the inuence of the dipolarity of the medium
by means of a solution of MS in ClB at temperatures over the
This journal is

the Owner Societies 2012

Fig. 2 Fluorescence spectra of MS dissolved in 2MB measured at

temperatures over the ranges 293113 K.

Phys. Chem. Chem. Phys., 2012, 14, 89038909

8905

Published on 11 January 2012. Downloaded by FAC DE QUIMICA on 24/04/2015 18:19:52.

View Article Online

Fig. 3 Normalized uorescence excitation spectra of MS dissolved in

2MB obtained monitoring light at 480 nm (black), at 400 nm (blue),
and 360 nm (red) in the sample at 113 K.

Fig. 5 Fluorescence spectra of MS dissolved in ClB measured at

temperatures over the ranges 343293 K.

(specically, at 480, 400 and 360 nm in the sample at 113 K).

The envelope for these spectra exhibits no vibronic structure;
also, the three are superimposable and so is the corresponding
absorption spectrum. Consequently, the emission above 360 nm
is produced by electronic excitation of a single MS form in the
ground state that must be MSn.
UVVis spectroscopy of MS in ClB
Fig. 4 shows the absorption spectra for an MS solution in ClB
at 293 to 113 K (i.e. in the liquid phase down to 253 K and in a
solid matrix below 133 K). The appearance of these spectra is
quite normal: decreasing the solution temperature substantially
increases their intensity and no appreciable change is observed
from the liquid phase to the solid phase. Interestingly, these
spectra contain an isosbestic point at ca. 326 below and above
which the intensity increases and decreases, respectively, as the
temperature is lowered; also, the envelopes exhibit no vibronic
structure at any temperature. Even more interestingly, the
absorption peak changes very little with temperature; this
suggests that the absorption maximum is independent of the
dipolarity of the medium, which changes dramatically from
293 (SdP = 0.560) to 113 K (SdP = 1.197).39
Fig. 5 and 6 show the emission spectra upon excitation at
290 nm for a solution of MS in ClB at temperatures over the
ranges 343293 and 293113 K, respectively. As can be seen,
the emission increases signicantly and its maximum is blueshifted as the temperature is lowered: from 458 nm at 343 K to
443 nm at 113 K. Below 153 K, the emission exhibits incipient

Fig. 6 Fluorescence spectra of MS dissolved in ClB measured at

temperatures over the ranges 293113 K.

vibronic structure, less marked than in 2MB (see Fig. 2), which is
completely inappreciable at higher temperatures. It should be
noted that this gain in vibronic structure in the emission spectra
below 153 K is absent from both the absorption spectra of Fig. 4
and the excitation spectra obtained at this temperature level.
As can be seen in Fig. 5, raising the temperature from 293 to
343 K causes the emission of MS with its onset at ca. 320 nmor,
possibly, at a lower wavelength by eect of self-absorption
and a maximum at ca. 336 nm to increase and that emission at
ca. 455 nm to signicantly decreaseso much so that the band
at 336 nm is stronger than that at 455 nm above 313 K.
Interestingly, these emissions exhibit an isosbestic point at
ca. 362 nm as well. No similar behaviour was observed in the
2MB solution, simply because this solvent boils at 30 1C, but
would have been seen in methylcyclohexane (Mecyh), which
boils at 101 1C. Fig. 7 shows the spectra for MS in Mecyh at
temperatures from 293 to 343 K, which coincide with those
obtained in ClB (Fig. 5) and are consistent with those
previously obtained by Ford et al.24 in cyclohexane.
UVVis spectroscopy of MS in SQ

Fig. 4 UV absorption bands of MS dissolved in ClB measured at

temperatures over the ranges 293113 K.

8906

Phys. Chem. Chem. Phys., 2012, 14, 89038909

Fig. 8 shows the emission spectra for a solution of MS in SQ at

temperatures from 293 to 193 K. It should borne in mind that
this solvent is liquid only above 235 K and also that it possesses
a high viscosity (34 cp at 293 K and about 19 000 p at 193 K42).
As can be seen, the spectra are essentially identical with those
obtained in 2MB (Fig. 2) and ClB (Fig. 6). Also, the band at
336 nm in the spectra for the compounds in ClB and Mecyh is
also present in those recorded in SQ but does not increase as the
temperature is raised from 293 to 343 K.
This journal is

the Owner Societies 2012

Published on 11 January 2012. Downloaded by FAC DE QUIMICA on 24/04/2015 18:19:52.

View Article Online

Fig. 7 Fluorescence spectra of MS dissolved in Mecyh measured at

temperatures over the ranges 343293 K.

Fig. 9 Normalized UV absorption band (black) and uorescence

excitation spectrum monitoring light at 340 nm (red) of MS dissolved
in ClB at 343 K.

Fig. 8 Fluorescence spectra of MS dissolved in SQ measured at

temperatures over the ranges 293193 K.

Fluorescence emissions of MS
In theory, MS can give three uorescence emissions in the
studied solvents. One has its onset at 320 nm and peaks at
336 nm that increases with increasing temperature except in
such a highly viscous medium as SQ. This emission can in
principle be assigned to the MSr form, which is present in very
low proportions and, as shown by Pimentel et al.,30 photochemically produced. This minor form exhibits an absorption band that
is very similar to that for the major form, MSn, but blue-shifted by
about 8 nm as calculated from the spectra of Fig. 9.
As can be seen from these emission spectra, see Fig. 2, 6 and
8, lowering the temperature shifts the onset to 340 nm in all
media. This result is especially important and was already
observed by Weller in the uorescence spectra for MS in
Mecyh at 25 and 180 1C (see Fig. 1 in ref. 1). To the authors
knowledge, nobody has to date noted that the onset of these
spectra is signicantly red-shifted at the lowest temperature
(from 320 nm at 25 1C to 333 nm at 180 1C). One can
therefore conclude that another uorescence emission starts at
337 nm and becomes increasingly apparentto the detriment of
the emission peaking at 336 nmas the solution temperature
is lowered.
Also, the broad uorescence band with its onset at 337 nm
starts to grow abruptly at 380 nm and becomes a peak at
about 450 nm and spanning the range to 600 nm. Judging by
the following facts, however, this band may be the result of
two dierent electronic transitions:
(a) The band exhibits incipient vibronic structure at very low
temperatures, but only its rst half (viz. from 337 to 450 nm,
This journal is

the Owner Societies 2012

Fig. 10 Fluorescence intensity of MS measured for the 460550 nm

spectral region vs. that at 340380 nm.

Fig. 2). This suggests that the rst uorescence transition may
start at 337 nm and end at about 450 nm.
(b) Further support for this spectral splitting can be
obtained by comparing the uorescence of two compounds
with an IMHB (MS and 7-hydroxy-1-indanone) in a durene
matrix at 4.2 K (see Fig. 5 and 9b in ref. 43).
(c) The relative emission intensity between the spectral zones
at 460550 nm and at 340380 nm is solvent- and temperaturedependent (see Fig. 10).
In principle, these two uorescence bands could be assigned
to UV emission from the MSn* form, directly produced by an
FC transition from MSn in the ground state and blue-shifted
with respect to MSt*the species produced by proton transfer
via the IMHB in MSn*. The following section discusses the
eciency of ESIPT between the two forms.
On the eciency of the ESIPT process via the IMHB in MS
The uorescence quantum yield of the zwitterionic form MSt*
at a given temperature, Ft(T), is a function of that with which
the uorescing form, MSn*, is formed and hence of the
eciency of the proton transfer involved in its ESIPT process,
Fpt(T), the rate constant for the uorescence emission, kt, and
the emission lifetime, tt(T):
Ft(T) = Fpt(T)  kt  tt(T)

(1)

from which the eciency of the ESIPT process can be

calculated:
Phys. Chem. Chem. Phys., 2012, 14, 89038909

8907

View Article Online

Table 1 Fluorescence quantum yields, Ft, and lifetimes, tt, of the

proton-transferred species, and quantum yields for the excited state
proton-transfer process eciency, Fpt, following eqn (2), for MS in
2MB solution and in ClB solution from 293 to 113 K

Published on 11 January 2012. Downloaded by FAC DE QUIMICA on 24/04/2015 18:19:52.

2MB

ClB
pt

T in K

Ft

tt

293
273
253
233
213
193
173
153
133
113

0.024
0.039
0.061
0.122
0.191
0.292
0.398
0.476
0.537
0.592

0.29
0.47
0.98
1.75
2.96
4.24
5.64
6.28
7.89
8.78

0.99
1.00
0.75
0.84
0.77
0.83
0.85
0.91
0.82
0.81

Ft

tt

Fpt

0.018
0.034
0.055
0.099
0.170
0.310
0.445
0.558
0.609
0.688

0.22
0.41
0.81
1.45
2.59
4.39
6.34
7.12
7.84
8.53

0.98
0.99
0.81
0.82
0.79
0.85
0.84
0.94
0.93
0.97

Fpt(T) = [Ft(T)/tt(T)]  kt

(2)

At low enough temperatures, radiationless processes in the

MSt* form in the solvents used in this work must be negligible
and deactivation of the sample be governed by its uorescing
process; therefore, if the proton transfer eciency, Fpt, were
unity, so would the uorescence quantum yield, Ft, and kt
could be readily calculated from the ratio Ft/tt.
Table 1 shows the quantum yields of the uorescence
spectra for MS in 2MB and ClB over the studied temperature
ranges. The yields were calculated from the areas under the
spectra of Fig. 2 and 6 over the range 390560 nm as corrected
for identical absorbance at 290 nm (the excitation wavelength
used) and identical refractive index squared in the solvents at
each temperature. The values were scaled with provision for
the quantum yield of MS in the solvents at 293 K relative to
quinine sulfate, which were 0.024 and 0.018, respectively. The
table also lists the lifetimes obtained.
Plots of the quantum yields in these solvents (Table 1)
against the corresponding temperatures were acceptably linear
(r was 0.985 in 2MB and 0.976 in ClB) and provided a
respective quantum yield at 0 K of 0.984  0.045 and 1.14 
0.07, i.e. two near-unity values which suggest that the quantum
yield of the proton transfer in MS, Fpt, must also be close to
unity in both solvents. Lifetime versus temperature plots were
also acceptably linear (r was 0.988 in 2MB and 0.980 in ClB)
and provided a respective lifetime estimate for MS at 0 K of
14.25  0.59 and 14.75  0.81 ns, consistent with the 12  2 ns
measured by Goodman and Brus19 in a neon matrix at 4.2 K
and the 12  0.5 ns obtained by Felker et al.34 under jet-cooled
conditions. This led us to use a kt 1 value of 1.2  10 9 s in
eqn (2). The Fpt(T) values thus obtained are also shown in
Table 1 and allow one to conclude that ESIPT in MS is highly
eective in both 2MB and ClB, and hence that no energy
barrier arises as the temperature is lowered. By contrast, the
double proton transfer in the 7-azaindole dimer is subject to a
barrier at low enough temperatures.40
Photophysical model for MS
The commonly accepted view that the dual uorescence of MS
is produced by two rotamers bearing an IMHB, MSn and
MSr, requires some clarication with a view to understanding
8908

Phys. Chem. Chem. Phys., 2012, 14, 89038909

some aspects of MS photophysics such as the inuence of

substituents.44,45
The so-called UV emission of MS is no doubt produced
by the MSr rotamer, electronic excitation of which causes it to
emit light from its FC structure since excitation fails to
increase the acidity of the hydroxyl group and basicity of the
ether group strongly enough to trigger a proton transfer in the
excited state. As a result, MSr gives an emission spectrum with
its onset slightly below 320 nm and maximum at ca. 336 nm,
and an excitation spectrum starting at 330 nm and peaking at
300 nm.
Based on theoretical calculations, Massaro and BlaistenBarojas36 recently showed the conversion of MSn into MSr to
be very dicult and excluded a role of the latter in the
photophysical behaviour of MS. However, Pimentel et al.30
previously showed that MS is photochemically converted into
MSr even by excitation at the wavelength of the 00 component of its rst p,p* singletin any case, the transformation
involves the rst triplet state of the compound, this may thus
be the origin of the presence of a small proportion of the
photorotamer MSr in an MS sample. We should also bear in
mind that raising the temperature of a solution of MS in
Mecyh or ClB above 293 K (Fig. 5 and 7) signicantly
increased the contribution of MSr as the temperature was
raised; however, eect was signicantly greater with MS in the
gas phase (see Fig. 5 in ref. 32) and undetectable in a solution
of MS in SQ. Based on the foregoing, the target phenomenon
is strongly dependent on the cavity available for the transformation, which, as stated above, involves twisting by 1801 an
ester group and is unfavoured by a large viscosity.
Electronic excitation of MSn produces an emission spanning
the wavelength range 337600 nm and peaking at about
450 nm. This situation is more complex than the previous
ones. As discussed above, this broad emission band is in fact
due to a combination of two dierent electronic transitions.
Thus, experimental evidence from the emission and excitation
spectra for MS support the assumption that this band is the
result of exciting the normal form of the compound, MSn.
Simultaneously to its generation, the FC form (MSn*) undergoes a dramatic increase in the acidity of the hydroxyl group
and basicity of the carbonyl group;20 because these two groups
form a hydrogen bond, this causes proton transfer between
them and the formation of MSt* as a result. As calculated
here, the eciency of the proton transfer from MSn* is near-unity
throughout the studied temperature range (293113 K) in both
2MB and ClB; this rules out the presence of an energy barrier and
its increase as the sample temperature is lowered. Therefore, the
potential energy surface (PES) governing this process cannot have
two minima, but rather a single minimumone that is located in
the zone for MSt* but broad enough to be reached by FC
transition from MSn and also for MSt* to reach FC in the
zone for MS. Clearly, these emissions must be weak but still
capable of generating the typical vibronic (staircase) structure
of MSn. Also, FC transitions from MSt* must be at signicantly lower energy levels and hardly able to generate vibronic
structure because the MSt structure is not a minimum in the
ground state PES.
Fig. 11 shows a photophysical scheme for MS consistent
with the previous arguments and also with recent theoretical
This journal is

the Owner Societies 2012

Published on 11 January 2012. Downloaded by FAC DE QUIMICA on 24/04/2015 18:19:52.

View Article Online

Fig. 11 Photophysical scheme proposed to summarize and explain the

photophysical behaviour of the normal molecular-structure of MS (MSn).

contributions such as that by Aquino et al.35 (The protontransfer curves for the p,p* state are very at. The assignment
of an energy minimum structure is arbitrary.) or that by
Massaro and Blaisten-Barojas,36 who suggested a two-minimum
PES from their calculated proton transfer barrier of 0.001 eV
(i.e. ca. 0.02 kcal mol 1). This behaviour should be seen as the
result of a strong increase in the acidity of the phenol group,
which fails to retain its proton in response to the strong
increase in the basicity of the carbonyl group, which captures
the proton. This diagram is only intended as a one-dimensional
plot of what is a more complex and multidimensional phenomenon. The proposed scheme allows one to understand the
process and hence, as shown elsewhere,46 to modulate the PES
for the process.

Conclusions
Methyl salicylate (MS) emits UV light with a maximum at
336 nm upon electronic excitation of a rotamer which is unable
to trigger an ESIPT process in its excited electronic state
despite it possessing an intramolecular hydrogen bond
(IMHB). This rotamer disappears as the temperature is lowered.
The emission of the normal form of MS spanning the wavelength
range 337600 nm can be seen as the combination of a weak
uorescence produced by deactivation of the 1(p,p*) state in the
compound to the energy minimum of MS in its ground state and
a stronger one due to FC transitions of the MSt form along a
repulsive segment of the curve for the ground state. Such an
interesting situation provides the means to adjust the behaviour
of MS in a proton transfer via its IMHB by increasing or
decreasing the acidity of its hydroxyl group.

Acknowledgements
This paper is dedicated to the memory of Albert Weller, in
appreciation of his pioneering research into such an interesting
topic as ESIPT processes.

References
1 A. Weller, Z. Elektrochem., 1956, 60, 1144.
2 Th. Forster, Z. Elektrochem. Angew. Phys. Chem., 1950, 54, 42 and 531.

This journal is

the Owner Societies 2012

3 A. Weller, Z. Elektrochem. Ber. Bunsenges. Phys. Chem., 1952,

56, 662; A. Weller, Z. Phys. Chem., 1955, 3, 238.
4 W. Urban, Diplom-Arbeit Stuttgart 1954.
5 A. Weller, Progress in Reaction Kinetics, ed. Porter, Pergamon
Press, London, 1961, vol. I, p. 187.
6 H. Beens, K. H. Grellmann, M. Gurr and A. Weller, Discuss.
Faraday Soc., 1965, 39, 183.
7 W. Klopfer, Adv. Photochem., 1977, 10, 311.
8 A. Khan and M. Kasha, Proc. Natl. Acad. Sci. U. S. A., 1983, 80, 1767.
9 A. Costela, F. Amat, J. Catalan, A. Dual, J. M. Figuera,
J. M. Munoz and A.U. Acuna, Opt. Commun., 1987, 64, 457.
10 J. Catalan, J. C. DelValle, F. Fabero and N.A. Garcia, Photochem.
Photobiol., 1995, 61, 118.
11 J. Keck, H. E. A. Kramer, H. Port, T. Hirsch, P. Fischer and
G. Rytz, J. Phys. Chem., 1996, 100, 14468.
12 J. Catalan, F. Fabero, M. S. Guijarro, R. M. Claramunt,
M. D. Santa-Maria, M. C. Foces-Foces, P. Hernandez-Cano,
J. Elguero and R. Sastre, J. Am. Chem. Soc., 1990, 112, 747.
13 Plastic packaging: interactions with food and pharmaceuticals, ed.
O. G. Piringer and A. L. Banner, Wiley VCH, Weinheim, 2008.
14 M. Chattoraj, B. A. King, G. U. Bublitz and S. G. Boxer,
Proc. Natl. Acad. Sci. U. S. A., 1996, 93, 8362.
15 R. Y. Tsien, Annu. Rev. Biochem., 1998, 67, 509.
16 W. H. Chen, Y. Xing and Y. Pang, Org. Lett., 2011, 13, 1362.
17 J. Catalan and J. I. Fernadez-Alonso, J. Mol. Struct., 1973, 18, 37.
18 J. Catalan and F. Tomas, Adv. Mol. Relax. Processes, 1976, 8, 87.
19 J. Goodman and L. E. Brus, J. Am. Chem. Soc., 1978, 100, 7472.
20 J. Catalan, J. Palomar and J. L. G. de Paz, J. Phys. Chem. A, 1997,
101, 7914.
21 W. Klopfer and G. Naundorf, J. Lumin., 1974, 8, 457.
22 K. Sandros, Acta Chem. Scand., Ser. A, 1976, 30, 761.
23 K. K. Smith and K. J. Kaufmann, J. Phys. Chem., 1978, 82, 2286.
24 D. Ford, P. J. Thistletwaite and G. J. Woolfe, Chem. Phys. Lett.,
1980, 69, 246.
25 A. U. Acuna, F. Amat-Guerri, J. Catalan and F. Gonza`lez-Tablas,
J. Phys. Chem., 1980, 84, 629.
26 F. Toribio, J. Catalan, F. Amat-Guerri and A. U. Acuna,
J. Phys. Chem., 1983, 87, 817.
27 L. A. Heimbrook, J. E. Kenny, B. R. Kohler and G. W. Scott,
J. Phys. Chem., 1983, 87, 280.
28 S. Melandri, B. M. Giuliabo, A. Maris, L. B. Favero, P. Ottaviani,
B. Velino and W. Caminati, J. Phys. Chem. A, 2007, 111, 9076.
29 R. D. Massaro, Y. Dai and E. Blaisten-Barojas, J. Phys. Chem. A,
2009, 113, 10385.
30 E. Orton, M. E. Morgan and G. C. Pimentel, J. Phys. Chem., 1990,
94, 7936.
31 J. D. Coe, B. G. Levine and T. J. Martinez, J. Phys. Chem. A, 2007,
111, 11303.
32 A. U. Acuna, J. Catalan and F. Toribio, J. Phys. Chem., 1981, 85, 241.
33 L. A. Heimbrook, J. E. Kenny, B. E. Kohler and G. W. Scott,
J. Chem. Phys., 1981, 75, 5201.
34 P. M. Felker, Wm. R. Lambert and A. H. Zewail, J. Chem. Phys.,
1982, 77, 1603.
35 A. J. A. Aquino, H. Lischka and C. Hattig, J. Phys. Chem. A, 2005,
109, 3201.
36 R. D. Massaro and E. Blaisten-Barojas, J. Chem. Phys., 2011,
135, 164306.
37 K. Law and J. Shoham, J. Phys. Chem., 1994, 98, 7472.
38 J. Catalan, J. Phys. Chem. B, 2009, 113, 5951.
39 J. Catalan, J. L. G. de Paz and C. Reichardt, J. Phys. Chem. A,
2010, 114, 6226.
40 J. Catalan, J. Phys. Chem. A, 2010, 114, 5666.
41 R. A. Velapoldi, J. Res. Natl. Bur. Stand. (U.S.), Sect. A, 1972,
76, 641.
42 R. D. Deegan, R. L. Leheny, N. Menon, S. R. Nagel and
D. C. Venerus, J. Phys. Chem. B, 1999, 103, 40664070.
43 T. Nishiya, S. Yamauchi, N. Hirota, M. Baba and I. Hanazaki,
J. Phys. Chem., 1986, 90, 5730.
44 A. U. Acuna, F. Toribio, F. Amat-Guerri and J. Catalan,
J. Photochem., 1985, 30, 339.
45 S. Nagaoka, A. Nakamura and U. Nagashima, J. Photochem.
Photobiol., A, 2002, 154, 23.
46 J. Catalan, publication in preparation.

Phys. Chem. Chem. Phys., 2012, 14, 89038909

8909