The six-membered intramolecular hydrogen bond position as a switch for inducing an excited state intramolecular proton transfer (ESIPT) in esters of o-hydroxynaphthoic acids

Article abstract of DOI:10.1021/jp992631p

The substituted naphthalene compounds investigated in this paper, i.e., methyl 2-hydroxy-3-naphthoate (MHN23), methyl l-hydroxy-2-naphthoate (MHN12), and methyl 2-hydroxy-l-naphthoate (MHN21), show a strong intramolecular hydrogen bond {-IMHB) in their ground electronic state. The relative position of the IMHB in the naphthalene skeleton acts as a switch and controls the yield of an excited state intramolecular proton transfer (ESIPT) process. As a matter of fact, only MHN23 exhibits a proton transfer (PT) emission and possesses a theoretically proved ESIPT mechanism. The role that the ESIPT mechanism plays on the photostability of the molecule MHN23 is unravelled by comparison with the model compounds methyl salicylate (MS), MHN12, and MHN21. On one hand, the low photoreaction quantum yield, N = 0.00015, and therefore the high photostability of MS, under direct ultraviolet (UV) irradiation, has been explained due to the photophysics of its proton transfer tautomer. On the other hand, (a) the two benzene-fused ring derivatives of methyl salicylate, MHN12 and MHN21, also possess a great photostability to UV radiation, and they do not support an ESIPT mechanism; and (b) although MHN23 exhibits an excited state proton transfer, its efficiency is only of 1.8%, and the photostability is 5 times larger than that of MS. As a result, the photostability of MHN23, MHN12, and MHN21 does not rely on the photophysics of their proton transfer tautomers but on the nonradiative dynamics of their respective normal tautomers. We present experimental evidence which supports the above-mentioned statement on the existence of distinctive nonradiative channels for the molecules MHN23, MHN12, and MHN21.

Full text of DOI:10.1021/jp992631p

J. Phys. Chem. A 1999, 103, 10921-10934
10921
The Six-Membered Intramolecular Hydrogen Bond Position as a Switch for Inducing an
Excited State Intramolecular Proton Transfer (ESIPT) in Esters of o-Hydroxynaphthoic
Acids
Javier Catal a´ n,* Juan C. del Valle, Jos e´ Palomar, Cristina D ´ı az, and Jos e´ L. G. de Paz
Departamento de Qu ´ı mica F ´ı sica Aplicada, UniVersidad Aut o´ noma de Madrid,
Cantoblanco, E-28049 Madrid, Spain
ReceiVed: July 28, 1999; In Final Form: October 13, 1999
The substituted naphthalene compounds investigated in this paper, i.e., methyl 2-hydroxy-3-naphthoate
(MHN23), methyl 1-hydroxy-2-naphthoate (MHN12), and methyl 2-hydroxy-1-naphthoate (MHN21), show
a strong intramolecular hydrogen bond (IMHB) in their ground electronic state. The relative position of the
IMHB in the naphthalene skeleton acts as a switch and controls the yield of an excited state intramolecular
proton transfer (ESIPT) process. As a matter of fact, only MHN23 exhibits a proton transfer (PT) emission
and possesses a theoretically proved ESIPT mechanism. The role that the ESIPT mechanism plays on the
photostability of the molecule MHN23 is unravelled by comparison with the model compounds methyl salicylate
(
MS), MHN12, and MHN21. On one hand, the low photoreaction quantum yield, Φ ) 0.00015, and therefore
r
the high photostability of MS, under direct ultraviolet (UV) irradiation, has been explained due to the
photophysics of its proton transfer tautomer. On the other hand, (a) the two benzene-fused ring derivatives
of methyl salicylate, MHN12 and MHN21, also possess a great photostability to UV radiation, and they do
not support an ESIPT mechanism; and (b) although MHN23 exhibits an excited state proton transfer, its
efficiency is only of 1.8%, and the photostability is 5 times larger than that of MS. As a result, the photostability
of MHN23, MHN12, and MHN21 does not rely on the photophysics of their proton transfer tautomers but
on the nonradiative dynamics of their respective normal tautomers. We present experimental evidence which
supports the above-mentioned statement on the existence of distinctive nonradiative channels for the molecules
MHN23, MHN12, and MHN21.
1
. Introduction
strong IMHB, but they do not undergo an ESIPT mechanism,
however, exhibit a high photostability to ultraviolet irradiation
The molecular skeleton of naphthalene exhibits two distinct
(vide infra). How could we explain this seeming controversy?
nonequivalent carbon pair positions, 1-2 and 2-3, to which a
six-membered intramolecular hydrogen bond (IMHB) ring can
be linked. This special feature does not appear in benzene
derivatives containing an IMHB ring, e.g., methyl salycilate
Thus, we aim at understanding the relationship which links the
photostability of intramolecular hydrogen bonded aromatic
molecules to an ESIPT mechanism.
(
MS). For the benzene molecule, all the carbon pair positions
Evidence on the existence of the intramolecular hydrogen
bonded MHN23 molecule has already been reported by Berg-
are equivalent. This observation allows us to study the photo-
physical influence that the relative position of the intramolecular
hydrogen bond ring has on the esters of o-hydroxynaphthoic
acids, benzene-fused ring analogues of MS (Scheme 1). Thus,
we aim at answering the following questions: does the IMHB
position in the naphthalene ring influence the occurrence of an
excited-state intramolecular proton transfer (ESIPT) process and
how does the IMHB position affect the photostability of the
molecule?
The molecules methyl 2-hydroxy-3-naphthoate (MHN23) and
methyl salicylate possess a strong IMHB in the ground electronic
state and yield a proton transfer emission with a low quantum
yield.1 In accordance with the general belief3-9 found in the
references, the great photostability shown (cf. Photostability
section) by MHN23 and MS relies on the photophysics of their
proton transfer tautomers (PT-tautomer, Scheme 2). On the other
hand, the molecules methyl 1-hydroxy-2-naphthoate (MHN12)
and methyl 2-hydroxy-1-naphthoate (MHN21) also present a
1
2
mann et al. The observation of a dual emission in the
fluorescence spectrum of MHN23 was first established by
1
3
Naboikin et al. Those emissions are a short-wavelength
fluorescence at ca. 410 nm and a long-wavelength emission
-
1
(exhibiting approximately a 10 000 cm shift from the first
14-16
absorption band), of the type described by Weller
for methyl
salicylate, at ca. 650 nm. The long-wavelength emission was
ascribed to the effects of the intramolecular hydrogen bond of
13
the molecule MHN23. Naboikin et al. ruled out the possibilities
of rationalizing the 650 nm emission band as it being produced
by dimer or excimer species. Finally, Woolfe and Thistlethwaite1
,2
ascribed the long-wavelength fluorescence to the excited state
proton transfer tautomer of MHN23 (PT-tautomer, Scheme 2).
Nevertheless, they placed the PT-tautomer fluorescence at 608
nm, which is not consistent with the results reported by Naboikin
1
3
et al. Our results show (vide infra) that, when the emission
spectrum is corrected for instrumental sensitivity, the wavelength
maximum shifts to ca. 650 nm, thereby matching the wavelength
*
To whom correspondence should be addressed. Fax: 34 91 397 4187.
1
3
E-mail: javier.catalan@uam.es.
value found by Naboikin et al.
1
0.1021/jp992631p CCC: $18.00 © 1999 American Chemical Society
Published on Web 11/25/1999

10922 J. Phys. Chem. A, Vol. 103, No. 50, 1999
Catal a´ n et al.
SCHEME 1
SCHEME 2
corresponding isomers, the N23-tautomer species, is predomi-
nant in the ground electronic state. The observation of only one
predominant isomer for the ground state is also common to
1
7
methyl salicylate and the salicylic acid family. Thus, we give
emphasis to the correction of the emission and excitation spectra
to acquiring a better understanding of the photophysical
phenomena.
1
Also, Woolfe and Thistlethwaite suggested that two ground
state conformers (Scheme 2, HB23 and NHB23-1) must exist
in order to explain the finding of a dual exponential decay time
for the short-wavelength emission band in cyclohexane and
methanol solutions.
Methyl salicylate and its naphthalene derivative MHN23 have
1
0,11
recently drawn attention
to interpreting the photophysical
behavior and ground state concealed isomers of methyl salicyl-
ate. Furthermore, the molecule salicylic acid and many of its
derivatives have been reported for being tribo-active materials
1
0,18-22
very useful in dry xerographic toners.
In addition, the photophysics of MHN12 and MHN21 has
received little interest in the literature; therefore, we are
presenting a thorough photophysical study of these molecules
in both cyclohexane solution and the gas phase.
For the first time, the photophysical influence of the IMHB
position in an aromatic molecule has been studied theoretically
(by means of hybrid density functional theory) and experimen-
tally. From this information, we unravel the important role that
the IMHB ring position plays in the excited state intramolecular
proton transfer for esthers of the o-hydroxynaphthoic acids.
According to Woolfe and Thistlethwaite1 the commonly
observed MHN23 fluorescence at ca. 410 nm exhibits a distinct
excitation spectrum compared to that of the long-wavelength
1
band, and neither resembles the UV absorption spectrum. We
2
. Experimental Section
show in this paper discrepancies with those results, because after
correcting the excitation spectra for instrumental sensitivity, the
excitation spectrum of the long-wavelength band definitely
matches the absorption spectrum. Thus, only one of the
Instrumentation. The absorption spectra for liquid samples
at 298 K were monitored on a Shimadzu UV-2100 spectrometer
using Suprasil cells of a 1 cm path length. The gas-phase

Proton Transfer in Esters of o-Hydroxynaphthoic Acids
J. Phys. Chem. A, Vol. 103, No. 50, 1999 10923
absorption spectra at 353 K were obtained with the aid of a
Cary 5 spectrometer using Suprasil cells of a 10 cm path length
at 353 K.
ones, as well as the results obtained with the highest post-HF
methods.
3
9-42
Geometry optimization and vibrational data for the structures
4
3,44
studied in this paper were computed at B3LYP
level using
Emission and excitation measurements were carried out on
an SLM Aminco-Bowman series 2 luminescence spectrometer
at the temperatures shown in the text or figure captions.
the 6-31G** basis set. For the sake of comparison between
experimental and calculated IR spectra, the calculated wave-
number values have been drawn to scale by multiplying by
The gas-phase emission spectra were recorded for samples
obtained by sublimating the compounds (MHN12, MHN21, and
MHN23) in a closed cell (1 cm path length) through which a
stream of 99.999% pure N2 was circulated for 30 min (prior to
the heating of the sample). All of the emission and excitation
spectra shown in this paper have been appropriately corrected
for instrumental sensitivity. The emission spectra were corrected
using as a standard a lamp of spectral irradiance (Optronic
Laboratories, Inc., OL 245 M).
4
5
0
.9613.
The evaluation of potential energy surfaces (PES) that
describe the proton motion entails a molecular structure
4
6
optimization. Ground state intramolecular proton transfer
curves (GSIPT) were calculated with the energies of the B3LYP/
6
-31G** fully optimized structures at fixed O-H distances over
the 0.9-1.6 Å range.
Information on the ESIPT mechanism was obtained by
calculating the Franck-Condon transition energies for the
B3LYP/6-31G** ground state structures at the CIS/6-31G**
Degassed samples were prepared with the aid of a vacuum
-
5
pump to 4 × 10 mbar using five freeze-pump-thaw cycles.
4
7
level. The Franck-Condon curves for the proton transfer
process were obtained by adding the CIS/6-31G** excitation
energies to the corresponding GSIPT curves. This modeling,
which does not include excited state optimization, has been
successfully applied to the ESIPT study of hydroxybenzoyl
compounds and displays a good agreement with the photo-
Quantum yields were calculated with the aid of the above-
mentioned instrument (SLM Aminco-Bowman) at 298 K on
freshly prepared samples with absorbances at the excitation
wavelength (293 nm) of ca. 0.05. The sample 2-aminopyridine
in 0.1 N H2SO4 was employed as a standard (Φf ) 0.66) to
measure the fluorescence quantum yields, which were corrected
taking into account the refractive indices of the solvents used.
48
physical evidence.
An energy increment value EPT is used for evaluating the
occurrence of the proton transfer process in the ground state.
This energy increment for the S ′ f S back proton transfer
Fluorescence lifetimes were determined with phase and
modulation measurements, relative to glycogen scattering solu-
tions by using an SLM 48000S spectrofluorimeter at frequencies
between 1 and 250 MHz. Decisions on the suitability of the
lifetimes rested on examination of the statistics of a fit (a plot
0
0
process is regarded as the difference between the energy at the
equilibrium distance of the N-tautomer and the energy of the
PT-tautomer at rOBH ) 1.6 Å
2
of the residual deviations with frequency) and the reduced ø
The strength of the intramolecular hydrogen bond (IMHB)
was estimated as the difference (EIMHB) between the energy of
the fully optimized molecular structures of the non-hydrogen-
bonded form (with a hydroxy group rotated 180°, i.e., NHB-1
in Scheme 1) and the energy of the N-tautomer molecule.
All the calculations have been performed with the Gaussian
2
values. In our experiments, the reduced ø values were all close
to unity. The detection limit of our intrument is ca. 30 ps.
Photostability at 298 K was determined by using a PTI
system, consisting of a horizontally mounted water-cooled lamp
housing for a 1000 W xenon lamp furnished with an f/4.0
elliptical reflector that concentrated the lamp intensity 4-fold.
Light reflected to the sample chamber was previously passed
through a water-cooled PTI 02-A002 infrared water filter and
then through a monochromator by which a wavelength was
selected for sample excitation. The photon flux reaching the
sample was measured with the Aberchrome 540.38 The samples
used were stirred continuously during the irradiation. Sample
degradation was monitored by UV spectroscopy on a Shimadzu
UV-2100 spectrometer and was expressed by means of the
quantum yield of photoreaction (Φr) as the ratio between the
number of moles converted or photodegradated and the einsteins
4
9
94 program.
Synthesis. Methyl 2-hydroxy-3-naphthoate (MHN23) was
purified by silica gel column chromatography using n-hexane/
ethyl acetate (1:1) as eluent. Methyl 1-hydroxy-2-naphthoate
(MHN12) and methyl 2-hydroxy-1-naphthoate (MHN21) were
obtained by refluxing for 1 day the corresponding acids with
methyl alcohol in the presence of sulfuric acid. Finally, MHN12
and MHN21 were purified by silica gel column chromatography
using n-hexane/ethyl acetate (80:20) and n-hexane/dichloro-
methane (80:20) as eluents, respectively.
(moles of photons) absorbed by the sample at the excitation
3. Hydrogen Bond Strength and Stability of the Rotamers
wavelength (340 nm for MHN12 and MHN21, 368 nm for
MHN23, and 308 nm for MS). All the photolysis experiments
were carried out in cyclohexane. None of the samples show
aggregation or non-Beer-Lambert behavior at the concentra-
tions used. The photoreaction quantum yields were appropriately
corrected taking into account the respective fluorescence
quantum yield of the compounds studied.
The theoretical stabilities (Table 1) indicate that the S0 state
intramolecular hydrogen bond of both molecules MHN12 and
MHN21 is stronger than that of methyl salicylate (MS). In
contrast, MS has a stronger S0 hydrogen bond than that of
MHN23. That fact is clearly developed by observing the
calculated distances of the six-membered IMHB ring of the
N-tautomers of these compounds (Table 2). Thus, a more
effective hydrogen bridge lengthens the O1′-H bond distance
The infrared (IR) data for the molecules MHN12, MHN21,
-
1
and MHN23 within the range 4000-450 cm were recorded
(
Scheme 1) and shortens the O1′‚‚‚O5′ and H‚‚‚O5′ distances of
on a Fourier transform Nicolet Impact 410 spectrometer at room
the MHN12 and MHN21 compounds with respect to those of
MS; while the converse distance variations apply to the MHN23
molecule. As can be seen in Table 2, there exists a resonance-
-
1
temperature with 200 scans and 1 cm in resolution. The IR
spectra were recorded in CCl4 and cyclohexane solutions.
50
Theoretical Calculations. Hybrid density functional methods
have been applied to several studies of molecular systems
containing an intramolecular hydrogen bond. These methods
predict molecular data that match the available experimental
assisted hydrogen bond mechanism that produces the shorten-
ing of the C2′-O1′ and C3′-C4′ distances (increased double bond
character) and the lengthening of the C4′-O5′ and C2′-C3′
distances (decreased double bond character). Thus, the proton

10924 J. Phys. Chem. A, Vol. 103, No. 50, 1999
Catal a´ n et al.
TABLE 1: Calculated Energy Values (in kcal/mol) at the B3LYP/6-31G** Level: IMHB Strength (EIMHB); Electronic Energy
Increments between the N-Tautomer and Both the HB (∆EHB-(N-tautomer)) and the NHB (∆ENHB-(N-tautomer)) Species; Free Energy
Increment between the N-Tautomer and the HB Species (∆GHB-(N-tautomer)); Energy Difference (EPT) between the N-Tautomer
(at the Equilibrium Distance) and the PT-Tautomer at a RO-H Distance of 1.6 Å
E
IMHB/∆ENHB-1-N-tautomer
∆EHB-N-tautomer
∆GHB-N-tautomer
∆ENHB-2-N-tautomer
E
PT
MHN21
MHN12
MS
15.6
15.4
12.3
11.5
3.4
4.3
3.8
3.7
2.8
4.1
3.6
3.2
13.3
14.7
11.7
10.7
13.1
13.0
17.6
21.5
MHN23
TABLE 2: Calculated Bond Distances (in Å) at the B3LYP/
-31G** Level for the Intramolecular Hydrogen Bond Ring
cf. Scheme 1) of the N-Tautomer (at the Equilibrium
Distance) and PT-Tautomer (RO-H ) 1.6 D)'
SCHEME 3
6
(
MHN21 MHN12 MS
MHN23
Normal Tautomer Bond Distances
O1′-H
0.997
1.334
1.415
1.471
1.240
1.591
2.497
0.992
1.338
1.405
1.464
1.237
1.679
2.576
0.987
1.342
1.42
1.469
1.234
1.72
0.985
1.346
1.439
1.473
1.232
1.73
O1′-C2′
C2′-C3′
C3′-C4′
C4′-O5′
O5′‚‚‚H
O1′‚‚‚O5′
2.606
2.614
Proton-Transfer Tautomer Bond Distances
O1′‚‚‚H
O1′-C2′
C2′-C3′
C3′-C4′
C4′-O5′
O5′-H
1.600
1.266
1.467
1.408
1.310
1.014
2.521
1.600
1.267
1.452
1.402
1.308
1.021
2.536
1.600
1.272
1.462
1.408
1.303
1.027
2.543
1.600
1.278
1.473
1.417
1.297
1.032
2.548
O1′..O5′
TABLE 3: Theoretical and Experimental Carbonyl and
Hydroxyl Stretching Frequency Values (in cm-1) for the
N-Tautomer and HB Species
ν(CdO)exp ν(CdO)theor ν(O-H)exp ν(O-H)theor
N21-tautomer
HB21
N12-tautomer
HB12
MSN-tautomer
MSHB
N23-tautomer
HB-23
1654
1728
1670
1730
1682
1731
1690
1736
1642
1719
1658
1734
1672
1735
1678
1737
3163
3434
3178
3468
3274
3506
3319
3526
3369
3397
3196
3450
3255
3481
Table 1 gathers the energy stabilities of the different rotamers
of MHN12, MHN21, and MHN23. Their relative stability
reveals that the ground-state equilibria N-tautomer (or HB) h
NHB (Scheme 1) should be shifted in favor of the species
containing an intramolecular hydrogen bond. The theoretical
free energies (Table 1) calculated at the B3LYP/6-31G** level
include electronic energies, unscaled zero-point energies, en-
thalpy temperature corrections, and absolute entropies at 298.15
K and 1 atm, derived by using the calculated harmonic
vibrational frequencies and the standard statistical thermody-
namic relationships. Thus, from the calculated free energies we
can estimate the equilibrium constants corresponding to the
different rotamers. For the ground state equilibrium N-tautomer
donor and acceptor groups are connected by a π-conjugated
double-bond system, whose principal effect is to make the
50
homonuclear IMHB stronger. By using infrared spectroscopy,
5
1,52
Hurnsberger
also drew the same conclusions.
The theoretical IR frequency values for the hydrogen bonded
N-tautomers resemble quite well the experimental IR spectra
in cyclohexane and CCl4 solutions. Table 3 illustrates the
excellent equivalence between the experimental IR carbonyl and
hydroxyl stretching frequencies and the respective theoretical
frequencies.
In assigning the molecular structures for the ground electronic
state (Schemes 1-3), we take into account the CdO and O-H
stretching frequencies (Table 3). Thus, the frequencies ν(CdO)
-
3
h HB, the equilibrium constants are as follow: 9.0 × 10
-3 -3
(MHN21); 1.0 × 10 (MHN12); 4.0 × 10 (MHN23). The
S0 chemical equilibria clearly lie in favor of the N-tautomer
species. On the basis of those equilibrium constants, the
population ratios in the gas phase for the N-tautomer vs HB
are ca. 111:1 for MHN21, 1000:1 for MHN12, and 250:1 for
MHN23.
(
experimental data) and ν(O-H) (theoretical data) appear shifted
to the red compared to those of the model compounds
1
(
2
increments of 82 cm for MHN21, 62 cm for MHN12, and
4
-1 50
-methylnaphthoate (νCdO ) 1724 cm ), 2-methylnaphthoate
-
1
50
-1
νCdO ) 1726 cm ), 1-naphthol (νO-H ) 3606 cm ), and
-
1
-naphthol (νO-H ) 3607 cm ). The carbonyl frequency shift
For the above-mentioned molecules, the IR spectra in CCl4
and cyclohexane solutions show a weak absorption band at ca.
-
1
-1
-1
-1
8 cm for MHN23, referred to the model compounds, indicate
1730 cm , which is also found for the molecule MS and was
5
4
the presence of strong intramolecular hydrogen bonds (in
sequence of decreasing strength), which according to Zador-
ozhnyi53 correspond to energy stabilities (EIMHB) of 10.2, 8.1,
and 5.2 kcal/mol, respectively.
tentatively assigned by Toribio et al. to the free carbonyl
stretching band of the HB isomer. Furthermore, our IR spectra
-1
indicate a small shoulder at ca. 3400 cm that may be
assignable to the O-H stretching frequency of the naphthol

Proton Transfer in Esters of o-Hydroxynaphthoic Acids
J. Phys. Chem. A, Vol. 103, No. 50, 1999 10925
TABLE 4: Electronic Energies (cm-1) and Oscillator
Strength Values of the First Excited Electronic Transition
Obtained at the CIS/6-31G**//B3LYP/6-31G** Level
MHN21
energy
MHN12
energy
MHN23
energy f
f
f
S
1
S
1
S
1
S
1
S
1
N-tautomer
PT-tautomer
HB
NHB-1
NHB-2
40 334 0.21 39 827 0.18 37 728 0.12
29 155 0.20
40 649 0.18 40 220 0.17 38 363 0.11
40 396 0.19 40 267 0.14 39 774 0.09
39 602 0.08
the six-membered IMHB ring (linked at the 1-2 or 2-3
positions) produces on the naphthalene system. As a matter of
fact, the first absorption band of anthracene (S0 f S1 band
maximum at 374.5 nm) is definitely 44.5 nm red shifted
compared to the first absorption band (S0 f S1 band maximum
Figure 1. Absorption spectra of the molecules MHN12 (dash line),
MHN21 (thick solid line), and MHN23 (thin solid line) in cyclohexane
solution at 298 K.
55
at 330 nm) of phenanthrene in n-hexane solution. In ac-
5
5,56
group, thus weakly bonded to the ether oxygen of the adjacent
ester susbtituent by means of an IMHB.54 It is noteworthy that
the wavenumber position and the intensity ratio of the IR bands
mentioned above do not depend on concentration and agree with
the calculated IR peaks of the respective rotamers (Table 2).
However, it is theoretically proved for methyl salicylate that
the frequencies for the free carbonyl groups of the rotamers
HB and NHB are almost identical. To check the assignment of
the free carbonyl frequency, the basicity of the solvent used
was increased, and thus an enhancement of the free carbonyl
cordance with Platt’s theory
(Scheme 4) the lowest excited
state transition is polarized in the short molecular axes for
1
1
anthracene ( La) and for phenanthrene ( Lb). Thus, by taking
into account our theoretical calculations, for MHN23 the
transition dipole moment of the S1 state is located in the short
1
molecular axis (being La as the one for anthracene) and the S1
transition dipole moments for MHN12 and MHN21 are forming
1
an angle of 33.9°, and 15.3°, respectively, referring to the Lb
axis of phenanthrene (Scheme 4). Thus, as an analogy of the
MHN21 and MHN12 molecules with phenanthrene and of
MHN23 with anthracene, a predominant intramolecular hydrogen-
bonded molecular structure (N-tautomer, Scheme 1) is expected
in hydrocarbon solution and the gas phase (cf. section 3). Also,
from the large absorption coefficient values, it is clear that the
first electronic transition is π,π* in nature for all the compounds.
-
1
-1
(
1730 cm ) and the hydroxy (3400 cm ) stretching band
intensities was monitored. This latter fact denotes that the
hydroxy group is intermolecularly bonded to a nucleophilic
heteroatom of the solvent, thereby forming NHB species.
Therefore, we also attribute the frequency localized at 1730
-
1
cm to the free carbonyl group.
On the other hand, from the theoretical IR intensities of the
carbonyl frequencies of the N-tautomers and HB species, in
conjunction with the baseline corrected experimental absorb-
ances of the 1730 cm-1 and 1650-1690 cm bands in
cyclohexane solution (Table 3), the relative presence of the
molecules N-tautomer vs HB is estimated to be 23:1 for
MHN21, 96:1 for MHN12, and 66:1 for MHN23. These
molecular ratios are consistent with the theoretically predicted
stability of the rotamers, thereby showing that the HB isomer
5. Emission in Cyclohexane Solution
Figures 2a and 3a depict the absorption and emission spectra
of the molecules MHN12 and MHN21, respectively. As
observed, the corresponding absorption and emission bands do
not show a very good mirror image symmetry, in part because
-1
-1
the absorption band half-width is 1400 and 1000 cm smaller
than the emission band half width for the molecules MHN12
and MHN21, respectively. However, the bandwidth anomalies
may imply that (a) other chemical structures can be contributing
to broaden the emission band of the two compounds above or
of MHN21 is especially stabilized compared to that of MHN12.
This stabilization is due to the peri effect3
0,32
of the hydrogen
(b) the S1 potential surface minimum suffers a major displace-
atom at position 8 (H-8) that effectively favors the twisting of
the esther group, because of the formation of a hydrogen bond
between the oxygen of the carbonyl group and the H-8 atom.
In summary, the theoretical and experimental IR data show
strong IMHBs, for the N-tautomer species, whose stabilization
energies are on the decrease for the following molecular
sequence MHN21, MHN12, and MHN23. The theoretical
equilibrium constants estimated predict that the N-tautomers are
also the main absorbing species in the S0 state, in accord with
the IR data in CCl4 and cyclohexane solutions.
ment from the S0 potential surface minimum, and thus a skeletal
distortion of the excited molecule is produced with a consequent
broadening of the emission band. As is shown in Figures 2b
and 3b, the excitation spectra of the respective emission bands
match the absorption spectra quite well. Thus, the emission
spectra come predominantly from the absorption spectra cor-
responding to mainly one of the chemical structures. Further-
more, for the reason that the excitation spectra obtained by
monitoring the whole range of the emission band resemble each
other, as well as the respective absorption spectra of MHN12
and MHN21, explanation (a) should be discarded. As a result,
the evident failure of Levshin’s law is explained by a displace-
ment between both potential energy curve minima of the S1
and So states.
4
. Absorption in Cyclohexane Solution
The absorption spectra of the molecules MHN12, MHN21,
and MHN23 in hydrocarbon solution are depicted in Figure 1.
As can be seen, the first absorption bands of the molecules
MHN12 and MHN21 are placed at approximately the same
wavelength region. However the first absorption band of
MHN23 is red shifted ca. 40 nm compared to the absorption
bands of MHN12 and MHN21. This is also consistent with the
theoretical S0 f S1 electronic energies gathered in Table 4. This
fact can be accounted for by the electronic modification that
The fluorescence quantum yields of the compounds MHN12
and MHN21 in cyclohexane are 0.19 and 0.04, respectively (on
excitation at 293 nm). No evidence of proton transfer emission
was monitored in the course of the experiments at dilute
solutions, neither at very concentrated solution nor at the crystal
phase for the MHN12 and MHN21 molecules. Recently, Tobita
1
1
et al. have reported the absence of the ESIPT mechanism in

10926 J. Phys. Chem. A, Vol. 103, No. 50, 1999
Catal a´ n et al.
SCHEME 4: S1 Transition Dipole Moments for the Molecules MHN12 (µ12), MHN21 (µ21), and MHN23 (µ23) Oriented
a
(r, â, γ) with Respect to the S1 Transition Dipole Moments (La, Lb) of Phenanthrene and Anthracene
a
IMHB ring position is indicated with hollow circles.
Figure 3. (A) Absorption (right) and fluorescence (left) spectra of
MHN21. (B) Absorption (solid line) and excitation (dash line) spectra
of MHN21. The spectra were recorded at 298 K in cyclohexane solution.
Figure 2. (A) Absorption (right) and fluorescence (left) spectra of
MHN12. (B) Absorption (solid line) and excitation (dash line) spectra
of MHN12. The experiments were carried out in cyclohexane at 298
K.
absorption spectra, the N-tautomer population in the ground state
must be present in a much larger contribution than that of HB23,
NHB23-1, and NHB23-2.
the molecule MHN12 by means of transient absorption in the
ground state, fluorometry, and time-resolved thermal lensing
(TRTL) measurements.
The MHN23 molecule exhibits two fluorescences (Figure 4)
6
. Lifetimes in Cyclohexane Solution
in cyclohexane solution with peaks at 400 nm and at ca. 650
nm and fluorescence quantum yields of 0.003 (this paper) and
For MHN12 on excitation at 335 nm, the lifetime modeling
3
2
0
.0007, respectively. On one hand, the 650 nm fluorescence
fits a decay time of 2.2 ns with a fractional contribution of 85%
and a second decay time of 0.7 ns. On excitation at 350 nm,
the same decay times are found but the fractional contribution
of the 2.2 ns decay time is 95%. From the theoretical
calculations (cf. section 3), a free energy of 4.1 kcal/mol is
predicted for the reaction N12-tautomer h HB12. Both mol-
ecules are in thermodynamic equilibrium in the ground state
with an equilibrium constant of 1.0 × 10 , which indicates
that the equilibrium should be almost entirely in favor of N12-
tautomer. Therefore, the absorption spectrum would bear a close
similarity to the excitation spectra, as is observed in Figure 2b.
ascribed to the PT-tautomer possesses an excitation spectrum
which resembles the MHN23 absorption spectrum (Figure 5).
On the other hand, the excitation spectrum of the short-
wavelength fluorescence is blue shifted with respect to the
absorption spectrum (Figure 5). This latter excitation spectrum
was recorded at different emission wavelengths from 390 to
-
3
5
00 nm without showing any significant difference in profile.
HB23, NHB23-1 and NHB23-2 (Scheme 2) could seemingly
account for the 410 nm emission band. In fact, since the
excitation spectra of the 650 nm emission band match the

Proton Transfer in Esters of o-Hydroxynaphthoic Acids
J. Phys. Chem. A, Vol. 103, No. 50, 1999 10927
-
1
s . The radiative constants (Kf ) Φf/τf) calculated from the
respective lifetimes and the fluorescence quantum yield (Φf )
7
-1
8 -1
0
0
.04) are 3.3 × 10 s (from 1.22 ns) and 1.2 × 10 s (from
.34 ns). As is observed, the radiative constant from the
absorption spectrum matches the one corresponding to the
lifetime of 1.22 ns, whereas the radiative constant corresponding
to the 0.34 lifetime is larger by about 1 order of magnitude. On
the grounds of the above reasonings we assign the 1.22 and
0
.34 ns decay times to the N21-tautomer and HB21 isomers,
respectively.
The decay time analysis for the short-wavelength emission
of MHN23 in cyclohexane solution seems to be complex. In
our measurements, we have employed a band-pass filter (Corion,
LS-500- T-S226) in order to monitor only the short-wavelength
fluorescence. The time-resolved analysis indicates that upon
excitation either at 340, 360, or 380 nm, the phase modulation
data reveal two decay times of 14.5 and 2.9 ns, with fractional
contributions of 90% and 10%, respectively. That aforemen-
tioned evidence indicates the existence of several species in the
S1 state emitting within approximately the same spectral region.
For the molecule MHN23 the electronic energy gap of NHB-1
with respect to HB or to N-tautomer is smaller than that between
the corresponding tautomers of MHN12 and MHN21 (Table
Figure 4. Dual fluorescence of MHN23 in cyclohexane at 298 K: (left)
short wavelength band; (right) proton transfer emission.
1
). In addition, there is an excellent matching between both the
excitation spectra of the proton transfer emission and the
absorption spectra of MHN23 (Figure 5), thereby meaning that
the ground state chemical equilibria are shifted to the N23-
tautomer. From our theoretical calculations, not only one but
two nonintramolecular hydrogen bonded species could be
present in the S0 state (NHB23-1 and NHB23-2, Scheme 2) in
view of the thermodynamic data, because the energy gap
between them is only of 0.8 kcal/mol, NHB23-2 being the most
stable conformer. However, the existence of NHB23-1 and
NHB23-2 is not in accordance with the results gathered in Table
Figure 5. Absorption spectrum (thin solid line), excitation spectrum
from the proton transfer emission (dash line), and excitation spectrum
from the short-wavelength emission (thick solid line) of the molecule
MHN23. All the experiments were measured in cyclohexane at 298 K.
In contrast, a very small equilibrium constant (k ) 5.1 × 10-12
)
1, in which these two species lie at 11.5 and 10.7 kcal/mol higher
is estimated from the electronic energy for the equilibrium N12-
tautomer h NHB12-1 (Table 1); thus, the contribution in
cyclohexane solution of NHB12-1 to the ground state population
should be neglected. As a result, since the excitation spectra of
the MHN12 fluorescence matches the absorption spectra and
the IR estimated ratio of the two feasible conformers, i.e., N12-
tautomer and HB12, is 96:1, respectively, the longer decay time
of 2.2 ns is assigned to the N12-tautomer and the shorter one
to the HB12 species. This interpretation is also supported by
the consistency between the radiative constants calculated from
than the N-tautomer. The calculated free energy values result
in agreement with the respective electronic energy values given
above. Since density functional calculations reproduce with great
accuracy the experimental free energy of other ground state
chemical reaction, as for example the ground state dimer-
formation reaction of 7-azaindole (J. Catal a´ n, to be published),
the assignment of the lifetimes to the NHB23-1 and NHB23-2
is quite uncertained. NHB23-1 and NHB23-2 are stabilized in
hydroxylic solvents by solute-solvent hydrogen bonding;
therefore, their contribution must be much less important in
cyclohexane solution. Thus, for MHN23 there are at least two
absorbing species (N23-tautomer, HB23) in the ground elec-
tronic state that could fluoresce within the short wavelength
emission band. On the other hand, from the absorption spectrum
7
-1
both the absorption spectrum of MHN12 (Kr ) 4.3 × 10 s )
according to the Strickler and Berg’s equation)57 and the ratio
(
between the fluorescence quantum yield and the lifetime (τf )
7
-1
2
.2 ns) of the N12-tautomer, Kf ) Φf/τf (8.6 × 10 s ).
For MHN21 on excitation at 320, 335, 350, and 360 nm, the
7
-1
in cyclohexane a radiative constant (Kr) of 2.1 × 10 s is
calculated, which is several orders of magnitude larger than
those radiative constants (Kf) estimated from the lifetimes and
lifetime modeling clearly yields two decay times of 0.34 and
.22 ns. Several considerations are made in order to assign these
1
decay times. As for the MHN21 molecule, there is also an
excellent matching between both the excitation and the absorp-
tion spectra (Figure 3b). In addition, the equilibrium between
N21-tautomer and NHB21 does not seem feasible in cyclohex-
ane solution since the equilibrium constant has such a small
5
-1
the fluorescence quantum yield (0.003), that is, 2.1 × 10 s
6
-1
(from 14.5 ns) and 1.0 × 10 s (from 2.9 ns). Thus, the
photophysical data from the short-wavelength emission are not
a consequence of the absorption spectrum and the main
absorbing species (N23-tautomer). As a result, the N23-tautomer
does not fluoresce within the short-wavelength band. At this
stage, two possibilities can account for the ca. 98% of the N23-
tautomer population generated in the S1 state upon direct
excitation: radiationless deactivation routes and the ESIPT
mechanism (cf., Proton Transfer section). As we estimate below
a proton transfer efficiency of 0.018, this process can be omitted
as the main channel of deactivation of the N23-tautomer species.
-
12
value (K ) 3.6 × 10 , also estimated from the electronic
-
3
energy in Table 1). An equilibrium constant of 9.0 × 10 for
the rotamers N21-tautomer and HB21 and the IR data (cf.
section 3) suggest that these two species are absorbing in the
ground electronic state and contribute to the fluorescence
spectra, though the ground state equilibrium is shifted to the
former rotamer. The radiative lifetime (28.2 ns) estimated from
7
the absorption spectra yields a radiative constant of 3.5 × 10

10928 J. Phys. Chem. A, Vol. 103, No. 50, 1999
Catal a´ n et al.
In conclusion, the lifetime of 14.5 ns is ascribed to the HB23
tautomer based upon IR data, its large fractional contribution
5
-1
of emission (90%), and its radiative constant of 2.1 × 10 s .
Consequently, the lifetime of 2.9 ns could be tentatively assigned
to a non-hydrogen bonded species (NHB). However, the species
NHB23-1 and NHB23-2 do not seem to be feasible candidates.
Aggregation effects can be precluded in these experiments
because the solutions used were sufficiently diluted. More
experiments are in progress in order to assign the 2.9 ns lifetime.
7
. Photostability
The quantum yields of photoreaction under direct ultraviolet
irradiation are for the first time reported for the molecules
MHN12, MHN21, MS, and MHN23 in cyclohexane solution
at 298 K (cf., Experimental section). Upon excitation at
approximately the wavelength maxima (λexc) of the first
absorption band, the following photoreaction quantum yields
-
4
are obtained: 6.1 × 10 for MHN21 (λexc ) 340 nm); 3.3 ×
-
4
-4
1
3
0
for MHN12 (λexc ) 340 nm); 1.5 × 10 for MS (λexc )
-
5
08 nm); 2.7 × 10 for MHN23 (λexc ) 368 nm). These
quantum yields are as low as those reported for the archetype
of photostability 2-(2′-hydroxy-5′-methylphenyl)benzotriazole
7
,34
(Tinuvin P).
All of the molecules above present strong
intramolecular hydrogen bonds, but it is surprising that the
molecule that contains the weakest one shows instead the highest
photostability (cf., sections 8 and 9).
8
. Gas-Phase Absorption and Emission
The gas-phase emission and absorption spectra (air equili-
brated) of the molecules MHN12, MHN21, and MHN23 are
depicted in Figures 6 and 7. The absorption spectra display
slightly more vibronic structure and are blue shifted (ca. 550
-
1
-1
-1
cm for MHN12, 450 cm for MHN21, and 400 cm for
MHN23) in comparison to those obtained in cyclohexane
solution. The red-shift dispersion found in cyclohexane solution
compared to the gas phase is due to the well-known polariz-
ability effect caused by the solvent.
Figure 6. Gas-phase (air-equilibrated) fluorescence spectra at 353 K
of the molecules MHN12, MHN21, and MHN23. A dual fluorescence
is also observed for MHN23, the short-wavelength emission (A) and
the proton transfer emission (B).
Excitation of the molecules MHN12 and MHN21 in the gas
phase gives rise to single fluorescence bands resembling those
obtained in cyclohexane solution, but blue shifted and showing
slightly less vibronic structure. It is important to highlight that
their excitation spectra exhibit much smaller half widths than
their respective gas-phase absorptions (Figure 7). This latter
observation is made more evident for the gas phase at vacuum
pressure conditions (degassed sample, without air), denoting that
the absence of collisional relaxation produces the loss of a large
part of the spectrum. These results are comparable to those
feasible explanation, at least in part, for the high photostability
demonstrated by MHN12, MHN21, and MHN23 in cyclohexane
solution.
9
. Proton Transfer
In this section we are concerned with two main points. On
2
reported for methyl salicylate at the same experimental
one hand, whether the ESIPT mechanism is responsible or not
for the excellent photostability to direct ultraviolet irradiation
exhibited by MHN23 in cyclohexane solution. On the other
hand, we should explain the observation of ESIPT mechanism
in the molecule MHN23 and its absence in MHN12 and
MHN21.
conditions, which were ascribed to the presence of an active
radiationless channel.The above-mentioned evidence (Figure 7)
demonstrates the presence of a radiationless deactivation channel
in the hydrogen bonded N-tautomers of the molecules MHN12
and MHN21.
A dual fluorescence is shown in parts a and b of Figure 6 for
MHN23, as is observed in cyclohexane solution, that is, the
long-wavelength emission corresponding to the PT-tautomer and
the short-wavelength emission arising from a mixture of emitting
species (cf. section 6). Figures 7 and 8 display the effect of
collisional relaxation on the long-wavelength and short-
wavelength emissions, respectively, and denote the presence of
radiationless deactivation pathways for the N23-tautomer and
HB23. This radiationless deactivation pathways could be a
Compounds containing an intramolecular hydrogen bond have
hitherto received great attention for being very good photo-
stabilizers because of their excellent photostability to ultraviolet
irradiation. The 2-(2′-hydroxyphenyl)benzotriazoles, methyl
salicylate and some of its derivatives exemplify this type of
molecules. Since the compounds above present low PT-tautomer
2
,4,7-9
fluorescence quantum yields, some authors
have invoked
the ESIPT mechanism to explain their great photostability. The
photostability would rely on a rapid radiationless deactivation

Proton Transfer in Esters of o-Hydroxynaphthoic Acids
J. Phys. Chem. A, Vol. 103, No. 50, 1999 10929
Tinuvin P) has been proposed as an archetype of photostability
due to the photophysics of its PT-tautomer. However, the
percentage of molecules transferring the proton in the excited
state has been estimated for Tinuvin P to be very low (ca.
2
8
0
.02%) in cyclohexane solution, and experiments in dimethyl
sulfoxide solutions show that its photostability is based upon
avoiding the generation of non-hydrogen-bonded species,3
which are in fact very reactive through the triplet state located
at ca. 500 nm. These chemically reactive triplet states are
quenched by another low lying triplet state at ca. 720 nm,58
which corresponds to the chemical structures with a nonper-
turbed IMHB ring.
0,34
The molecules 3-hydroxyflavone, salicylamide,25 sodium
2
4
2
6
26,27
salicylate, and the 2-(o-hydroxyphenyl)benzimidazoles.
exhibit quite a distinct photophysical behavior compared to that
of the 2-(2′-hydroxyphenyl)benzotriazoles. They present high
proton transfer fluorescence quantum yields, and the ESIPT
mechanism is recalled to explain their high gain coefficient
values, since they are used as dye lasers.
Some other molecules as for example the phenylperim-
3
5,36
32
idines
and 1′-hydroxy-2′-acetonaphthone present a strong
intramolecular hydrogen bond, but do not yield an ESIPT
mechanism, and they are even more photostable than Tinuvin
P.
On the other hand, good photostabilizers, as for example the
2
-hydroxybenzophenones and methylsalicylates, exhibit a high
percentage of excited state proton transfer molecules. For
instance, the methyl salicylate molecule has ca. 95% of the
5
4
population undergoing a proton phototransfer in cyclohexane
solution. Therefore, the great photostability (Φr ) 0.00015) of
methyl salicylate is definitely assignable to the photophysics
of its PT-tautomer.
Figure 7. (1) Gas-phase (air-equilibrated) absorption spectra recorded
using a 10 cm pathlength. (2) Gas-phase (air-equilibrated) excitation
spectra of the compounds MHN12 and MHN21 and gas-phase (air
equilibrated) excitation spectrum from the proton transfer emission of
MHN23. (3) Gas-phase (without air, at a pressure of 4 × 10 mbar)
excitation spectra of the molecules MHN12 and MHN21. All the spectra
were measured at 353 K.
Therefore, it is a crucial point to estimate the percentage of
MHN23 PT-tautomer species generated. Thus, the fluorescence
quantum yield of the proton transfer tautomer is given by the
C
following expression, ΦF ) ΦTPΦF , that is, the product of the
-
5
proton transfer efficiency (ΦTP) and the intrinsic fluorescence
C
quantum yield of the PT-tautomer (ΦF ). As it is known, the
quantum yield ΦFC is equal to the product τF kF , and since
C
C
the PT-tautomer does not absorb in the ground state, the
C
fluorescence radiative constant (kF ) must be estimated from
C
lifetime (τF ) measurements at low temperature, assuming it is
independent of temperature. Therefore, the final expression for
the proton transfer efficiency is given as follows, ΦTP )
C
C
[
ΦF(room temperature)τF (low temperature)]/τF (room temper-
ature). From the literature (for degassed samples), the lifetimes
at room temperature (at 298 K in methylcyclohexane) and low
temperature (at 77 K in methylcyclohexane) are 0.23 and 0.59
10
ns, respectively. Taking into account an intrinsic fluorescence
1
0
quantum yield of 0.007, the estimated percentage of MHN23
molecules evolving to the excited state PT-tautomer is as low
as 1.8%. Thus, we can conclude that the great photostability
(Φr ) 0.000027) shown by MHN23 in cyclohexane solution is
not due to the photophysics of its PT-tautomer but to the
photophysics of the N23-tautomer. For the sake of consistency,
we have employed all the data from ref 10. However, there exist
quantitative discrepancies of those data with ours. Our fluores-
cence quantum yields for MHN23 are much smaller than those
given in ref 10. For instance, we have measured a proton transfer
fluorescence quantum yield which is 10 times smaller than that
given in ref 10. As for our measurements, all the fluorescence
quantum yields have been measured 10 times, they being within
a 15% stardard deviation. Also, the lifetime of the red emission
Figure 8. Gas-phase excitation spectra at 353 K from the short-
wavelength emission of MHN23, (2) Air-equilibrated and (3) without
-
5
air at a pressure of 4 × 10 mbar.
of the quinonoid form (PT-tautomer), which is caused by the
approximation of the electronic states implicated in the intramo-
lecular proton transfer emission.
The 2-(2′-hydroxyphenyl)benzotriazoles constitute a most
important class of photoprotecting agents, for instance, the 5′-
methyl derivative (2-(2′-hydroxy-5′-methylphenyl)benzotriazole,
1
from MHN23 was measured as 60 ps vs a lifetime of 230 ps

10930 J. Phys. Chem. A, Vol. 103, No. 50, 1999
Catal a´ n et al.
Figure 9. GSIPT curve (S
MHN12. The stretching O-H coordinate is given in angstroms.
0
), and ESIPT Franck-Condon curves of
Figure 10. GSIPT curve (S
MHN21. The stretching O-H coordinate is given in angstroms.
0
) and ESIPT Franck-Condon curves of
1
reported in ref 10. However, the detection limit of our instrument
The 1(π,π*) potential energy curves of MHN12 and MHN21
(
e30 ps) is much better than that given in ref 10 (e200 ps).
Thus, if we take into account our data in conjunction with the
.59 ns lifetime for the 77 K experiment from ref 10, a lower
excited state proton transfer efficiency of 0.007 is obtained
0.7%).
Figures 9, 10, and 11 show the proton transfer potential
exhibit a single minimum at the equilibrium distance of the
N-tautomer and therefore intramolecular proton transfer is
hindered. Instead of that, MHN23 presents a 1(π,π*) state with
1
0
double-well, and an exothermal behavior of the potential curve
of the PT-tautomer, which explains the occurrence of the proton
transfer fluorescence observed in cyclohexane solution. The
calculations suggest that enol-keto (N-tautomer f PT-tau-
tomer) tautomerization is completed after ESIPT. The calculated
energy barrier for the ESIPT process of MHN23 is around 1.5
kcal/mol. We can conclude from the aforementioned theoretical
data that the ESIPT mechanism is only energetically favorable
in the molecule MHN23. The existence of two potential energy
wells could enable the existence of two fluorescences, that is,
from the N23-tautomer and the PT-tautomer (both corresponding
to hydrogen bonded species), but the former emission has not
been detected (cf. section 6).
(
energy curves either in the ground or excited states for the
molecules MHN12, MHN21, and MHN23, respectively. As can
be seen, the calculations at the level B3LYP/6-31G** provide
ground state intramolecular proton transfer (GSIPT) curves with
a single minimum at the equilibrium O-H distance of the
N-tautomer. The GSIPT curves from the PT-tautomers are
repulsive which implies that the proton transfer fluorescence
does not show a vibronic structure, as is observed in Figures 4
and 6b for MHN23. The GSIPT curves for MHN12 and MHN21
48
are markedly stabilized with respect to that of MHN23 or MS.
This markedly stabilized GSIPT curve for the proton transferred
form, also found for the molecule 1-hydroxy-2-acetonaphthone,
was explained as a result of the higher polarizability of the
naphthalene residue.46 Thus, this polarizability interaction might
be approximately the same for the three naphthalene type
compounds. However, the ground state energy increment value
A decisive insight on the different profiles of the GSIPT and
ESIPT curves can be acquired by analyzing the ground-state
molecular orbitals HOMO and LUMO (Figures 12, 13, and 14)
along the proton-transfer coordinate of the molecules studied.
5
9
The HOMO-LUMO transition chiefly describes the first
singlet excited states of all the molecules studied, in other words,
1
(cf., Experimental section, EPT, Table 1) for the molecule
the [1(π,π*) ] state is characterized by exciting one electron
MHN23 (21.5 kcal/mol) is much larger than that of MHN12
from the π HOMO (highest occupied molecular orbital) to the
π* LUMO (lowest unoccupied molecular orbital). Both HOMO
and LUMO are of π-type, however their phases are not equal
for the molecules MHN12, MHN21, and MHN23. Indeed, the
HOMO orbital of the N12-tautomer correlates with the highest
occupied π orbital of 1-naphthol, and the HOMO orbitals of
N21-tautomer and N23-tautomer correlate with that of 2-naph-
(13.1 kcal/mol) and MHN21 (13.0 kcal/mol).
The differences of relative EPT values of MHN21, MHN12,
and MHN23 with respect to MS cannot be rationalized due to
structural constraints, since all of them undergo similar bond
distance changes because of the proton tranfer process and
possess the characteristic hydrogen bond distances O1′‚‚‚O5′ and
O5′‚‚‚H (Table 2).
6
0
thol. The HOMO orbital on the IMHB system of MHN12

Proton Transfer in Esters of o-Hydroxynaphthoic Acids
J. Phys. Chem. A, Vol. 103, No. 50, 1999 10931
O1′ atom, and also shifted to the C3′-C4′ bond. As is shown in
Figure 13, the MHN21 compound has an HOMO projection
on the IMHB system with similar characteristics to that of
HOMO orbital of MHN12 (Figure 12). In contrast, HOMO is
distinctly projected on the IMHB system of MHN23 (Figure
1
3
2
4), principally because of the existence of a node at position
of the naphthalene ring (C3′, Scheme 1), as it is found for the
60
-naphthol HOMO orbital. Furthermore, there is no electronic
density contribution over the ester group of MHN23. Proton
transfer in the ground state produces an HOMO orbital of
MHN23 with less electronic density projection over C3′-C4′
and O5′ than that of the HOMO orbital of the molecules MHN12
and MHN21.
In conclusion, for the ground state of MHN23, the lower
conjugation through the IMHB system and the lower basicity
of carbonyl group weakens its IMHB strength, that together with
a less effective electronic transfer along the proton transfer
coordinate make the GSIPT process nonviable. The opposite
effects are found for the MHN12 and MHN21 molecules,
thereby strengthening the IMHB system and stabilizing the
GSIPT potential energy curves with respect to that of MS.
Nevertheless, this stabilization is not sufficient for producing a
GSIPT process.
On the other hand, the LUMO is a π* orbital for the three
compounds studied that also presents C2′O1′ and C4′O5′ anti-
bonding character, as the HOMO electronic map, but with a
C3′C4′ bonding character and some C2′C3′ antibonding contribu-
tion on the IMHB system (Figures 12-14). Weighty differences
are also found between MHN23, and both MHN12 and MHN21.
The LUMO orbital of N23-tautomer (Figure 14) possesses a
high projection on the O5′ atom and a null projection on the O1′
atom (in the opposite way to the HOMO), whereas for MHN12
Figure 11. GSIPT curve (S
MHN23. The stretching O-H coordinate is given in angstroms.
0
) and ESIPT Franck-Condon curves of
(Figure 12) and MHN21 (Figure 13), the LUMO orbital presents
(
Figure 12) is primarily of C2′C3′[C4′] bonding character, in
slightly higher projection on the O5′ atom and lower projection
on O1′ than that for the respective oxygen atoms of the HOMO
orbital. After tautomerization, the LUMO orbital of the PT-
tautomer of MHN23 still exhibits a large electronic density
projection on the O5′ atom, and almost null projection on the
O1′ atom, thereby favoring the proton transfer process. In
addition to the C2′O1′ and C4′O5′ antibonding contribution. Both
oxygen atoms have a bonding character, with a larger HOMO
projection over the hydroxyl oxygen (O1′). For MHN12 (Figure
1
2), the electronic rearragements induced by ground state proton
transfer lead to an HOMO projection still very large over the
0 1
Figure 12. Two-dimensional plots of the HOMO and LUMO orbitals, which are involved in the electronic excitation from S to S of the N-tautomer
and the PT-tautomer (RO-H ) 1.6 Å) of MHN12. The HOMO and LUMO drawings are evaluated 0.3 Å above the molecular plane. The nodal
planes are represented by dashed lines, and the contour lines are drawn every 0.03 au.

10932 J. Phys. Chem. A, Vol. 103, No. 50, 1999
Catal a´ n et al.
Figure 13. Two-dimensional plots of the HOMO and LUMO orbitals, which are involved in the electronic excitation from S
0
to S
1
of the N-tautomer
and the PT-tautomer (RO-H ) 1.6 Å) of MHN21. The HOMO and LUMO drawings are evaluated 0.3 Å above the molecular plane. The nodal
planes are represented by dashed lines, and the contour lines are drawn every 0.03 au.
0 1
Figure 14. Two-dimensional plots of the HOMO and LUMO orbitals, which are involved in the electronic excitation from S to S of the N-tautomer
and the PT-tautomer (RO-H ) 1.6 Å) of MHN23. The HOMO and LUMO drawings are evaluated 0.3 Å above the molecular plane. The nodal
planes are represented by dashed lines, and the contour lines are drawn every 0.03 au.
contrast, for the proton-transfer process of MHN12 and MHN21,
the LUMO orbital still shows a large electronic density
projection on the O1′ atom, and thus the proton transfer is
strongly hindered.
To sum up, the ESIPT process is driven from the N-tautomer
to the PT-tautomer of MHN23 because of the drastic changes
in acidity and basicity shown by the functional groups upon
excitation and the associated changes in electrostatic attraction

Proton Transfer in Esters of o-Hydroxynaphthoic Acids
J. Phys. Chem. A, Vol. 103, No. 50, 1999 10933
References and Notes
of the bridging hydrogen. Since this latter evidence is absent
for the S1 state of MHN12 and MHN21, an ESIPT mechanism
is precluded for these molecules.
(
1) Woolfe, G. J.; Thistlethwaite, P. J. J. Am. Chem. Soc. 1981, 103,
949.
2) Acu n˜ a, A. U.; Catal a´ n, J.; Toribio, F. J. Phys. Chem. 1981, 85,
3
2
(
41.
Also, we should call attention to the fact that for the three
molecular systems studied the second excited singlet electronic
(
3) Heller, H. J. Eur. Polym. J. Suppl. 1969, 105.
1
state [2(π,π*) ] does not go through a proton transfer process
(4) Williams, D. L.; Heller, A. J. Phys. Chem. 1970, 74, 4473.
(
Figures 9-11). This evidence is being currently tested in our
(5) Heller, H. J.; Blattman, H. R. Pure Appl. Chem. 1972, 30, 145.
(
(
(
6) Heller, H. J.; Blattman, H. R. Pure Appl. Chem. 1973, 36, 141.
7) Otterstedt, J. E. A. J. Phys. Chem. 1973, 58, 5716.
8) Kl o¨ pffer, W. In AdVances in Photochemistry; Pitts, F. J. N., Jr.,
laboratory.
1
0. Conclusions
Hammond, S. S., Gollnick, K., Eds.; Wiley: New York, 1977; Vol. 10, p
311 and references therein.
The photophysical, IR, and theoretical data reveal that the
(9) Werner, T. J. Phys. Chem. 1979, 83, 320.
10) Law, K.-Y.; Shoham, J. J. Phys. Chem., 1994, 98, 3114.
(
compounds studied possess predominant absorbing hydrogen
bonded N-tautomer species in the ground electronic state. The
intramolecular hydrogen bond strength for the N-tautomers
according to IR and theoretical data is on the decrease for the
following molecular sequence MHN21, MHN12, and MHN23.
Though MHN23 presents the weakest intramolecular hydrogen
bond, upon excitation it is definitely the only molecule which
undergoes an ESIPT mechanism. For the molecules MHN12,
MHN21, and MHN23, the relative position of the intramolecular
hydrogen bond on the naphthalene ring (i.e., the 1-2, 2-1,
and 2-3 linkings) determines both its strength and the occur-
rence of an ESIPT mechanism. The intramolecular hydrogen
bond strength is, on one hand, explained by the conjugation of
the IMHB ring with the naphthalene ring; thus, the stronger
conjugation the greater the IMHB strength is. As is theoretically
demonstrated, the electronic density distributions indicate an
IMHB ring, which is much more conjugated with the naphtha-
lene ring for the molecules MHN12 and MHN21 than for
MHN23. On the other hand, the position of the IMHB in the
molecular skeleton influences greatly the electronic transfer
produced upon excitation between the carbonyl and hydroxyl
groups (involved in the intramolecular hydrogen bond), which
increases their basicity and acidity, respectively, thereby favoring
the ESIPT mechanism for the molecule MHN23. In contrast,
no evidence of such definite electronic transfer within the IMHB
ring is shown by the molecules MHN12 and MHN21.
(
11) Tobita, S.; Yamamoto, M.; Kurahayashi, N.; Tsukagoshi, R.;
Nakamura, Y.; Shizuka, H. J. Phys. Chem. A 1998, 102, 5206.
12) Bergmann, E. D.; Hirshberg, Y.; Pinchas, S. J. Chem. Soc., 1950,
2351.
(
(13) Naboikin, U. V.; Zadorozhnyi, B. A.; Pavlova, E. N. Opt. Spectrosc.
(
Eng. Transl.) 1959, 6, 312.
(14) Weller, A. Naturwiss. 1955, 42, 175.
(15) Weller, A. Zs. Elektrochem. 1956, 60 (9), 1144.
(
16) Weller, A. Prog. React. Kinet. 1961, 1, 189.
(17) Woolfe, G. J.; Thistlethwaite, P. J. J. Am. Chem. Soc. 1980, 102,
6
917.
18) Nomura, Y.; Ide, N.; Ohtaki, K.; Tomita, M.; Tosaka, H.; Nanya,
T.; Orihara, M.; Chiba, S.; Inoue, S.; Asahina, Y.; Fushimi, H. U.S. Patent
(
4
,762,763, 1988.
19) Kawagishi, Y.; Narita, S.; Kiriu, T.; Uomoto, K. U.S. Patent
(
4
,656,112, 1987.
(20) Kiriu, T.; Arakawa, M. U.S. Patent 4,845,003, 1985.
(21) Kiuchi, T.; Maki, I. U.S. Patent 4,206,064, 1980.
(
22) Hashimoto, K.; Maruta, M.; Soyama, H.; Ishii, Y. U.S. Patent
,767,688, 1988.
23) Khan, A. U.; Kasha, M. Proc. Natl. Acad. Sci. U.S.A. 1983, 80,
1767.
4
(
(24) Chou, P.; McMorrow, D.; Aartsma, T. J.; Kasha, M. J. Phys. Chem.
1
984, 88, 4596.
(
25) Acu n˜ a, A. U.; Costela, A.; Mu n˜ oz, J. M. J. Phys. Chem. 1986, 90,
2807.
(26) Acu n˜ a, A. U.; Amat, F.; Catal a´ n, J.; Costela, A.; Figuera, J. M.;
Mu n˜ oz, J. M. Chem. Phys. Lett. 1986, 132 (6), 567.
27) Costela, A.; Amat, F.; Catal a´ n, J.; Douhal, A.; Figuera, J. M.;
(
Mu n˜ oz, J. M.; Acu n˜ a, A. U. Opt. Commun. 1987, 64 (5), 459.
(28) Catal a´ n, J.; Fabero, F.; Guijarro M. S.; Claramunt, R. M.; Santa
Mar ´ı a, M. D.; Foces-Foces, M. C.; Hern a´ ndez-Cano, F.; Elguero, J.; Sastre,
R. J. Am. Chem. Soc. 1990, 112, 747.
All the molecules studied exhibit a great photostability to
direct ultraviolet irradiation, comparable to the photostability
exhibited by some of the most renowned standards (e.g., methyl
salicylate and Tinuvin P). Furthermore, the quantum yield of
photoreaction of MHN23 is times smaller than that of the
molecules MHN12, MHN21, and methylsalicylate. The low S1
f S1′ excited state proton transfer efficiency of 0.018 discards
the ESIPT mechanism as the principal cause of photostability
for the molecule MHN23. As a result, the great photostability
of MHN23 does not rely on the photophysics of the PT-tautomer
but on the photophysics of the N23-tautomer. Nonradiative
mechanisms of the N23-tautomer are to be studied. The triplet
state photophysics may have an important role on this point,
and thus data on intersystem crossing (ISC) yields would be
important to understand the deactivation pathways in these
molecules. Also, for MHN23, an ESIPT process from the triplet
state manifold offers a new pathway which remains under
research.
(
29) Sastre, R.; Catalina, F.; Mateo, J. L.; Claramunt, R. M.; Santa Mar ´ı a,
M. D.; Catal a´ n, J. J. Polym. Sci.: Polym. Chem. 1990, 28, 3661.
30) Catal a´ n, J.; P e´ rez, P.; Fabero, F.; Wilshire, J. F. K.; Claramunt, R.
M.; Elguero, J. J. Am. Chem. Soc. 1992, 114, 964.
31) Catal a´ n, J.; Fabero, F.; Claramunt, R. M.; Santa Mar ´ı a, M. D.;
(
(
Foces-Foces, M. C.; Hern a´ ndez-Cano, F.; Mart ´ı nez-Ripoll, M.; Elguero,
J.; Sastre, R. J. Am. Chem. Soc. 1992, 114, 5039.
(
32) Catal a´ n, J.; del Valle, J. C. J. Am. Chem. Soc. 1993, 115, 4321.
(33) Catal a´ n, J.; del Valle, J. C.; Claramunt, R. M.; Santa Mar ´ı a, M.
D.; Bobosik, V.; Mocelo, R.; Elguero, J. J. Org. Chem. 1995, 60, 3427.
(34) Catal a´ n, J.; del Valle, J. C.; Fabero, F.; Garc ´ı a, N. Photochem.
Photobiol. 1995, 61, 118.
(
35) Catal a´ n, J.; del Valle, J. C.; Claramunt, R. M.; Sanz, D.; Dotor, J.
J. Lumin. 1996, 68, 165.
36) del Valle, J. C.; Catal a´ n, J.; Foces-Foces, M. C.; Llamas-Saiz, A.
L.; Elguero, J.; Sanz, D.; Dotor, J.; Claramunt, R. M. J. Lumin. 1997, 75,
7.
(
1
(37) Catal a´ n, J.; de Paz, J. L. G.; Torres, M. R.; Tornero, J. D. J. Chem.
Soc., Faraday Trans. 1997, 93, 1691.
(38) Heller, H. G.; Langan, J. R. J. Chem. Soc., Perkin Trans. 2, 1981,
3
41.
(
(
(
39) Barone, V.; Adamo, C. Int. J. Quantum Chem. 1997, 61, 429.
40) Barone, V.; Adamo, C. J. Chem. Phys. 1996, 105, 11007.
41) Lampert, H.; Mikenda, W.; Karpfen, A. J. Phys. Chem. 1997, 100,
Acknowledgment. We are greatly indebted to DGICYT of
Spain (Project PB93-0280) for financial support. One of us (J.P.)
acknowledges with thanks the granting of an F.P.I. scholarship
by the Ministry of Education and Science of Spain. C. D ´ı az
acknowledges Comunidad Aut o´ noma de Madrid for a postdoc-
toral fellowship. We are also grateful to Centro de Computaci o´ n
Cientifica de la Facultad de Ciencias (CCCFC) for CPU
facilities.
22254.
(42) Hass, K. C.; Schemeider, W. F.; Est e´ vez, C. M.; Bach, R. D. Chem.
Phys. Lett. 1996, 263, 414.
(43) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(44) Lee, C.; Yang, W.; Parr, R. G. Phys. ReV. 1988, B37, 785.
(45) Wong, M. W. Chem. Phys. Lett. 1996, 256, 391.
(
46) Catal a´ n, J.; Palomar, J.; De Paz, J. L. G. Chem. Phys. Lett. 1997,
69, 151.
47) Foresman, J. B.; Head-Gordon, M.; Pople, J. A. J. Phys. Chem.
1992, 96, 135.
2
(

10934 J. Phys. Chem. A, Vol. 103, No. 50, 1999
Catal a´ n et al.
(
48) Catal a´ n, J.; Palomar, J.; De Paz, J. L. G. J. Phys. Chem. A 1997,
01, 7914.
49) Frisch, M. J.; Trucks, G. W.; Head-Gordon, M.; Gill, P.; Wong,
(54) Toribio, F.; Catal a´ n, J.; Amat-Guerri, F.; Acu n˜ a, A. U. J. Phys.
Chem. 1983, 87, 817.
(55) Berlman, I. D. Handbook of Fluorescence Spectra of Aromatic
Molecules; Academic: New York, 1965.
1
(
M. W.; Foresman, J. B.; Johnson, B. J.; Schlegel, H. B.; Robb, M. A.;
Repogle, E. S.; Gomperts, R.; Andr e´ , J. L.; Raghavachari, K.; Binkley, J.
S.; Gonzalez, C.; Martin, R. L.; Fox, D. J.; Defrees, D. J.; Baker, J.; Stewart,
J. P.; Pople, J. A. Gaussian 94, revision D.1; Gaussian Inc.: Pittsburgh
PA, 1996.
(56) Platt, J. R. J. Chem. Phys. 1949, 17, 484.
(57) Strickler, S. J.; Berg, R. A. J. Chem. Phys. 1962, 37, 814.
(58) Catal a´ n, J. Chem. Phys. Lett. 1998, 297, 549 and results to be
published.
(59) The HOMO- f LUMO+1 excitation represents the second major
contribution (ca. 15%) for all of the S1 transitions of the molecules studied,
except for the one corresponding to the PT-tautomer of MHN23, which is
described exclusively as an HOMO f LUMO excitation.
(60) Utsunomiya, C.; Kobayashi, T. Nagakura, S. Bull. Chem. Soc. Jpn.
1975, 86, 1852.
1
(50) Gilli, G.; Belluci, F.; Ferreti, V.; Bertolasi, V. J. Am. Chem. Soc.
1
989, 111, 1023.
(
51) Hunsberger, I. M. J. Am. Chem. Soc. 1950, 72, 5626.
(52) Porte, A. L.; Gutowsky, H. S.; Hunsberger, I. M. J. Am. Chem.
Soc. 1960, 82, 5057.
53) Zadorohnyi, B. A.; Ischenko, I. K. Opt. Spectrosc. 1965, 19, 306.
(