Molecular structure, conformational preferences and vibrational analysis of 2-hydroxystyrene: A computational and spectroscopic research

Article abstract of DOI:10.1016/j.chemphys.2010.06.014

The molecular structure of 2-hydroxy-styrene has been investigated at DFT (B3LYP, mPW1PW91) and MP2 levels with an assortment of Pople's and Dunning's basis sets within the isolated molecule approximation. The presence of intramolecular hydrogen bonds has

Full text of DOI:10.1016/j.chemphys.2010.06.014

Chemical Physics 374 (2010) 62–76
Contents lists available at ScienceDirect
Chemical Physics
journal homepage: www.elsevier.com/locate/chemphys
Molecular structure, conformational preferences and vibrational analysis of
2-hydroxystyrene: A computational and spectroscopic research
Gregorio García a, Amparo Navarro a, José Manuel Granadino-Roldán a, Andrés Garzón a, Tomás Peña Ruiz a,
a
b
b
c
c
}
Maria Paz Fernández-Liencres , Manuel Melguizo , Antonio Peñas , Gábor Pongor , János Eori ,
Manuel Fernández-Gómez a,
*
a Department of Physical and Analytical Chemistry, University of Jaén, Campus Las Lagunillas, E23071 Jaén, Spain
b Department of Inorganic and Organic Chemistry, University of Jaén, Campus Las Lagunillas, E23071 Jaén, Spain
c Institute of Chemistry, Eötvös Loránd University, Pázmány Péter sétány 1/A, 1117 Budapest, Hungary
a r t i c l e i n f o
a b s t r a c t
Article history:
The molecular structure of 2-hydroxy-styrene has been investigated at DFT (B3LYP, mPW1PW91) and
MP2 levels with an assortment of Pople’s and Dunning’s basis sets within the isolated molecule approx-
imation. The presence of intramolecular hydrogen bonds has been theoretically characterized through a
topological analysis of the electron density according to the Atom-In-Molecules, AIM, theory. The confor-
mational equilibrium has been pursued by means of an analysis of the hydroxyl–phenyl and vinyl–
phenyl internal rotation barriers. This analysis also allowed getting an insight into the effects governing
the torsion barriers and the preferred conformations. A twofold scheme has been used for this goal, i.e.
the total electronic energy changes and the natural bonding orbital, NBO, schemes. The vibrational spec-
Received 7 April 2010
In final form 9 June 2010
Available online 13 June 2010
Keywords:
2-Hydroxystyrene
Molecular structure
Conformational dynamics
Topological analysis
Infrared and Raman spectra
Vibrational analysis
*
trum was recorded and then calculated at DFT-B3LYP/6-31G and cc-pVTZ levels. Two scaling methods,
SQMFF and linear scaling, have been applied on the theoretical spectrum in order to analyse the exper-
imental one. The results point out that at least three different conformers coexist at room temperature.
Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction
correlation between the frequency of O–H stretching, Hꢀ ꢀ ꢀ
p(vinyl)
and C–C(vinyl) distances with the O–H bond length in order to con-
clude about the possible existence and strength of hydrogen bonds.
They concluded that in the case of hydroxystyrenes the intramo-
lecular H-bonds are very weak.
Ortho-substituted phenols are model systems for studying
intramolecular O–Hꢀ ꢀ ꢀY hydrogen bonds and have been widely
investigated both experimental and theoretically (see [1] and ref-
erences therein). These compounds possess a common proton do-
nor (phenolic O–H group), whereas the proton acceptor may vary
within a wide range. Pauling [2], already in 1936, suggested for
many monosubstituted phenols the existence of cis and trans ori-
entations of the hydroxyl group with respect to the substituent
as well as the partial double-bond character of the C–O bond. This
double-bond character should tend the whole molecule to be
planar.
In the present work, we focus our attention on the ortho-vinyl-
phenol for which a non-negligible coupling between the so-called
large amplitude motions may arise due to the proximity of the vi-
nyl and hydroxyl moieties. This coupling may affect the conforma-
tional equilibrium and, ultimately, the vibrational spectrum.
Simperler et al. [1] studied systematically different kinds of pro-
ton acceptors at DFT-B3LYP level of theory in combination with 6-
31G(d,p), 6-311G(d,p), and DZVP basis sets, and they established a
On the other hand, the vibrational spectra of ortho-substituted
phenols have been the object of several works due to the number
of O–H stretching bands could be indicative of the number of iso-
mers present for a particular substituent [3].
To our knowledge, neither experimental structure nor vibra-
tional spectra of the title compound have been studied so far.
The first study on the molecular structure of 2-hydroxystyrene
was performed by Dubnikova and Lifshitz [4], who studied only
one conformer at B3LYP/cc-pVDZ level. At the time of writing this
report, we have been warned about a recently appeared paper
dealing with a conformational analysis and interpretation of
m(O–H) stretching bands of ortho-vinyl-phenol at B3LYP/cc-pVTZ
and 6-311G(d) levels wherein the authors conclude about the exis-
tence of three pairs of enantiomers plus one another flat rotamer in
gas phase and the stabilizing effect of a non-polar solvent, e.g. CCl4,
onto those conformers without intramolecular hydrogen bond. Be-
sides, these authors conclude on the small significance of such
intramolecular hydrogen bonds on the conformational mobility
of this system [5].
* Corresponding author. Tel.: +34 953 212148; fax: +34 953 212940.
E-mail address: mfg@ujaen.es (M. Fernández-Gómez).
0301-0104/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved.
doi:10.1016/j.chemphys.2010.06.014

G. García et al. / Chemical Physics 374 (2010) 62–76
63
The work is organized as follows:
The theoretical, vibrational spectrum was obtained at B3LYP/6-
31G* and B3LYP/cc-pVTZ levels. In the first case, the initial Carte-
sian force constants obtained from Gaussian 03 were appropriately
transformed into the set of natural internal coordinates as defined
by Fogarasi et al. [21]. The transformed set of the quadratic force
constants was then scaled by means of the SQM methodology
[22] and the SCALE3 [23] code. In the second case, a linear scaling
was used to fit both the calculated and the observed infrared spec-
trum [24].
1. Molecular structure and internal rotation barriers for 2-
hydroxystyrene are investigated in order to check the perfor-
mance of the different theoretical methods essayed. We have
performed full geometry optimizations with MP2 and DFT
(B3LYP and mPW1PW91) methods in combination with differ-
ent basis sets (6-31G*, 6-311++G**, cc-pVDZ and cc-pVTZ).
2. The study of barriers to internal rotation reveals important
information about the nature of bonding. The nature of the
effects governing the torsion barriers has been analysed using
two different partitioning schemes, i.e. the total electronic
energy [6] and the natural bond orbital, NBO [7]. The individual
torsion barriers have been deconvoluted by a sixfold Fourier-
type expansion.
3. The coupling between both rotors has been analysed on the
basis of a two-dimensional grid of 381 points constructed by
varying the vinyl– and the hydroxyl–phenyl dihedral angles in
the 0–180° range. The resulting energy surface vs. the torsion
dihedrals was then fitted to a suitable two-dimensional trigo-
nometric expansion assuming two coupled twofold-tops.
4. Atoms-in-Molecules theory (AIM) [8] has been applied to detect
the possible intramolecular hydrogen bond present in the
molecular system which could justify the preference for one
conformation among others.
5. The synthesis of the title compound was followed and the infra-
red (neat liquid and Cl4C solution), Raman (neat liquid) and 1H
NMR (CDCl3, solution) spectra were recorded and analysed at
room temperature. A multiple scaled quantum mechanical
force field, SQMFF, following Pulay’s methodology been
performed in order to assign the 3N-6 fundamental modes com-
paring the experimental vibrational spectra and those calcu-
lated for the different rotamers. The vibrational analysis may
shed light on the number of conformers present as the stretch-
ing and torsion of the hydroxyl group may be affected by the
relative orientation of the other substituents [3]. With this
aim, we have also performed a linear scaling of the frequencies
at B3LYP/cc-pVTZ in the isolated molecule approximation in the
high frequency zone.
2.2. Synthesis
2-Hydroxystyrene was synthesized according to the procedure
described in the literature [25], which consists in a Wittig’s
reaction on the salicylaldehyde (see Scheme 1). During the trans-
formation steps, all the operations were performed under argon
atmosphere. A suspension of MePPh3Br (8.583 g, 24 mmol) in dry
THF (50 ml) was added to solid t-BuOK (2.693 g, 24 mmol) and
the resulting mixture was stirred for 2 h at room temperature.
The reaction mixture was then cooled at ꢁ78 °C, salicylaldehyde
(1.16 ml, 10.914 mol) was slowly added, and the mixture was stir-
red overnight at room temperature. A saturated ammonium chlo-
ride solution was added and the aqueous phase was extracted
with diethyl ether (3ꢂ). The organic phase was dried over MgSO4,
filtered and concentrated under reduced pressure. The compound
was purified by flash chromatography on silica–gel (30% AcOEt,
70% hexane) yielding a yellow oil (1.121 g, 90.6% yield).
2.3. Spectral measurements
The infrared spectrum of the neat liquid was recorded using a
FT-IR Perkin–Elmer Spectrometer model 1760X, with KBr optics
and FR-DGTS detectors with 1 cmꢁ1 resolution and using windows
spacers of different thickness. Besides, the infrared spectrum of a
solution of 2-hydroxystyrene in CCl4 was recorded at room tem-
perature and 1 cmꢁ1 resolution by using a FT-IR Bruker Vector 22
spectrometer (CsI optics, DTGS detector). Raman spectra of the
neat liquid were finally recorded at 1 cmꢁ1 resolution using a FT-
Raman Bruker RFS100/S spectrometer equipped with an Nd-YAG
laser emitting at 1064 nm with a power of 500 mW and a liquid
N2 cooled Ge detector.
2. Experimental section
The 1H NMR spectrum was obtained from a solution of 2-
hydroxystyrene (20 mg) in CDCl3 (0.60 ml), at 299 K in a Bruker
Avance instrument operating at 400.13 MHz. Pulses of 30° were
employed with recycle delay of 3.31 s. Sixteen transients were
accumulated. This record was used to assess the identity and pur-
ity of the sample [26].
2.1. Computational details
The title compound was first modelled with the program Gauss-
View [9] and the suite of programs Gaussian 03 [10] was used to
carry out the ab initio MP2 [11] and DFT calculations running on
an ia64HP rx 2600 server. Pople’s 6-31G and 6-311++G [12] ba-
sis sets and Dunning’s correlation consistent basis sets, cc-pVDZ
[13] and cc-pVTZ [14] were used. In all cases, the molecular struc-
ture was optimized without imposing any restriction.
*
**
3. Results and discussion
3.1. Molecular structure
As exchange functional, Becke’s hybrid exchange B3 [15] was
used in conjunction with the non-local correlation Lee–Yang–Parr
(LYP) functional [16,17]. Also, the mPW1PW91 model was used
as a modified Perdew–Wang exchange functional and Perdew–
Wang 91 correlation functional [18]. The nature of the stationary
points was checked through the eigenvalues of the Hessian matrix.
The individual rotational barriers of the vinyl and the hydroxyl
substituents at B3LYP, mPW1PW91 and MP2 levels were calcu-
lated using the 6-31G* basis set whereas the coupling between ro-
tors was analysed with the help of a two-dimensional mesh
calculated at B3LYP and MP2 with 6-31G*. Natural bond orbital
NBO calculations and atoms-in-molecules AIM analysis have been
performed with MP2, B3LYP and mPW1PW91 methods in combi-
According to the initial, relative values of the vinyl– and the
hydroxyl–phenyl dihedral angles, four different rotamers can be
defined which are interconnected by rotation around single
r
O
OH
OH
MePPh3Br, tBuOK, THF
*
**
nation with 6-31G and 6-311++G basis sets using the NBO 3.1
code [19] and the AIM2000 package [20], respectively.
Scheme 1. Synthesis of 2-hidroxystyrene.

64
G. García et al. / Chemical Physics 374 (2010) 62–76
Fig. 1. Initial rotamers of 2-hydroxystyrene showing atom numbering.
bonds (see Fig. 1). No consideration about the existence of enantio-
mer pairs has been paid. The initial rotamers were afterwards opti-
mized in the isolated molecule approximation using the MP2,
spect to conformer 1. The order of stability becomes 1 > 2 > 3 > 4
*
for DFT methods and 1 > 2 > 4 > 3 for MP2/(6-31G , cc-pVTZ). Fi-
nally, for MP2/6-311++G** and cc-pVTZ the order becomes
1 > 4 > 2 > 3 and 1 > 2 ꢄ 4 > 3, respectively. In any case, the energy
gaps decrease as the basis set size increases, except for conformer 3
when using B3LYP and Dunning’s basis sets. No correction due to
zero-point energy has been considered. This result matches that
by Berdysehv et al. [5] although these authors also found that on
the basis of the electronic total energy and the Gibb’s free energy
the nature of the global minimum moved to conformer 2.
*
B3LYP and mPW1PW91 methods in combination with 6-31G , 6-
311++G**, cc-pVDZ and cc-pVTZ basis sets. As a result, MP2 with
*
**
Pople’s 6-31G and 6-311++G basis sets predicts the existence
of four stable rotamers, namely conformers 1–4 while DFT meth-
ods with the same basis sets only render conformers 1, 3 and 4.
They all, but conformer 3, show a non-planar conformation at
*
DFT (regardless the functional and the basis set) and MP2/6-31G
levels, the deviation from planarity being a function not only of
the conformer itself but also of the level of theory.
On the basis of the electronic energy, conformer 1 is predicted
to be the most stable whatever the level of theory with the C4–
In Table 1S (supplementary material) the full list of optimized,
geometrical parameters for the different conformers can be seen.
The influence of the method in the final, optimized structure
has been assessed through the root mean square, RMS, deviation
between the results for a given level of theory/basis set couple re-
spect to the B3LYP values adopted as the reference (see Fig. 1S). For
conformers 1, 3 and 4 the greatest sensitivity is shown by the O16–
C3–C1–C2 torsional angle,
a
ꢃ 35° and the H17–O16–C2–C5 tor-
sional angle, b ꢃ 10°. Table 1 lists the electronic energy differences
as well as the fractional Boltzmann population at 298.15 K with re-

G. García et al. / Chemical Physics 374 (2010) 62–76
65
Table 1
Finally, as regards the C2–C1–C3–C4 dihedral angle, the values
for conformers quasi trans/cis for 2-fluorostyrene become 149.2°/
0.0°, respectively, while for conformers 2–4/1–3 of 2-hydroxysty-
rene they are 147.9–134.7°/46.6–15.7°, respectively.
Energy differences (kcal molꢁ1) and fractional population (%) respect to conformer 1.
Level of theory
B3LYP
Conformer
2
3
4
*
6-31G
0.78 (26.95)
0.21 (70.26)
0.45 (46.94)
0.45 (46.94)
1.23 (12.65)
0.53 (41.03)
1.19 (13.53)
0.72 (29.81)
3.2. Internal rotation barriers
**
6-311++G
cc-pVDZ
cc-pVTZ
0.41 (50.20)
0.28 (62.46)
The nature and magnitude of the torsional barriers as well as
the position of the energy minima as a function of the level of the-
ory have been analysed in a one-dimensional approach. Default
convergence criteria as implemented in Gaussian 03 have been
used. The calculations have been made in a relaxed manner in such
mPW1PW91
MP2
6-31G*
6-311++G**
cc-pVDZ
1.10 (15.74)
0.61 (35.87)
0.84 (24.37)
0.87 (23.17)
1.29 (11.44)
0.68 (31.88)
1.26 (12.02)
0.89 (22.40)
0.68 (31.88)
0.60 (36.47)
cc-pVTZ
*
6-31G
0.96 (19.92)
0.57 (38.36)
1.14 (14.71)
0.76 (27.87)
1.94 (3.85)
1.39 (9.67)
1.27 (11.83)
1.14 (14.71)
1.07 (16.55)
0.43(48.54)
1.14 (14.71)
0.88 (22.78)
a way that a(b) is allowed to vary in the range 0–180° while b(a) is
**
6-311++G
cc-pVDZ
cc-pVTZ
kept equal to the value(s) for the optimized (opt) structure(s). No
further restriction has been imposed onto any other molecular
parameter. The energy profiles have been constructed respect to
{
a
(b) = 0°; b(
a
)
opt} and then fitted to a sixth-order Fourier expan-
C2 bond when using mPW1PW91 (deviation respect to B3LYP val-
ues ca. 0.010 Å, on average) while when using MP2, this is shown
by the C4–C3 bond (ca. 0.007 Å on average). However, for con-
former 2 the O2-H17 bond is that appearing to be by far the most
sensitive (ca. 0.430 Å deviation) both for mPW1PW91 and MP2
with cc-pVTZ.
The bond angle that most deviates from the B3LYP values is C4-
C3-C1 for conformers 1 (ca. 2.7°) and 3 (ca. 2.4°) both at MP2 level
whereas for conformers 3 and 4 it is the bond angle H17–O16–C2
that shows such a high sensitivity (ꢃ1.3° and ꢃ1.6° of deviation)
both at MP2 level.
Finally, the maximum sensitivity of dihedrals is shown by C4–
C3–C1–C2 and H9–C3–C1–C2 whatever the conformer considered.
For conformers 1 and 4, the maximum deviation amounts on aver-
age to ꢃ10.8° at MP2 level with Pople’s basis sets while for con-
formers 2 and 3, however, the maximum deviation amounts to
ꢃ18° at mPW1PW91/Dunning’s basis sets and ꢃ14.5° at MP2/6-
311++G** level.
The influence of the basis sets has been pursued through the
RMS deviations for bond distances, bond angles and dihedral an-
gles as obtained from using Pople’s and Dunning’s basis sets (see
Fig. 2S). The maximum deviation for bond lengths is shown by
H17–O2 bond for conformer 2 at B3LYP with Dunning’s basis sets,
i.e. 0.423 Å while for bond angles the maximum sensitivity is
shown by H17–O16–C2 bond angle for conformer 4 at MP2/Dun-
ning’s basis sets, i.e. 1.1°. Finally, the most sensitive dihedrals turn
out to be C4–C3–C1–C2 and H9–C3–C1–C2 for conformer 2 at
B3LYP/Dunning’s basis sets and conformer 3 at MP2/Pople’s basis
sets level, for which the RMS deviation amounts on average to
14.2°.
sion like:
X
1
2
D
E ¼ Vð
u
Þ ¼
VN ½1 ꢁ cosðiN
uÞꢅ;
u
¼
a; b
ð1Þ
i
i
where N is the symmetry number (1 in this case).
All the barriers were studied at B3LYP, mPW1PW91 and MP2
levels with 6-31G* basis set. Fig. 2 shows the theoretical, one-
dimensional potential energy surfaces (PESs) linking the different
conformers while Table 2S shows the values for Vi parameters ob-
tained from Eq. (1). To our knowledge, no experimental values for
such potential parameters are known to be compared with the the-
oretical ones.
The fitting is acceptable for all barriers, except for barrier a,
which could point towards a non-negligible coupling between both
rotors. Note that for the barriers a and c (vinyl rotating) the main
parameters are {Vi, i = 1–4} while for barriers b and d, the main
parameters are {V1, V2}.
According to Carlson and Fateley [3] the parameters V1 and V2
are needed to describe the hydroxyl rotation barrier in o-substi-
tuted phenols being the former an indicative of the existence of a
hydrogen bond contact between both substituents.
The magnitude of the different torsional barriers,
DEi (i = 1–6),
for each level of theory appears in Table 2. As can be seen, the rota-
tional barrier for the hydroxyl group turns out to be higher than
that for the vinyl group regardless the level of theory. The barriers,
measured from the most relaxed conformer, lie in the ranges ca.
3.32–5.27 kcal molꢁ1 and 0.65–3.67 kcal molꢁ1, for hydroxyl and
vinyl moieties, respectively while the hydroxyl rotational barrier
for phenol keeps lower than 3.5 kcal molꢁ1 [28]. Oki and Iwamura
[28] studied some ortho-allyl-phenols (as 2-hydroxystyrene), and
according to them the oxygen’s lone pairs of the hydroxyl are con-
jugated with the benzene ring, in such a manner that the O16–C2
bond would have a partial character of double bond what would
make the hydroxyl rotational barrier to increase and the hydro-
xyl–phenyl moieties to tend to be coplanar. This is actually the case
of conformers 2 and 3.
In order to analyse the effect of the ortho-substitution in sty-
rene, we have compared some structural parameters of 2-hydroxy-
styrene with 2-fluorostyrene at MP2/6-311++G** level [27]. Thus,
the values for C10–C3 and C3–C4 bond lengths turn out to be ca.
0
1.47 ÅA and 1.35 ÅA, respectively, both for conformers quasi-trans
and cis of 2-fluorostyrene and any of the conformers of 2-hydroxy-
styrene, pointing this way that they are rather insensitive to both
the nature and the relative orientation between the ortho substitu-
ent and the vinyl moiety.
3.3. Coupling between rotors
The comparison between C2–C1–C3 and C1–C3–C4 bond an-
gles, however, does show some substituent and relative orientation
dependence. Thus, the values for the quasi-trans conformer of 2-
fluorostyrene, 120.6° and 127.7°, compare with those for conform-
ers 2 and 4 of 2-hydroxystyrene, i.e. 120.0° and 123.8°, respec-
tively. However, when comparing the same bond angles between
the cis conformer of 2-fluorostyrene and the conformers 1 and 3
of 2-hydroxystyrene, the greatest difference lies on the C2–C1–
C3 bond angle for the conformer 1, 125.6° vs. 119.9° and on the
C1–C3–C4 bond angle for the conformer 3, 128.9° vs.125.0°.
Fig. 3S shows the evolution of the vinyl– and the hydroxyl–phe-
nyl dihedral angles for the different barriers and conformers. As a
result, the coupling between rotors for barrier c is almost negligible
while for the remaining barriers the correlation follows the order
b > a > d. In order to study the coupling between rotors we con-
structed a grid of 361 points by varying
a and b, in steps of 10°
in the range 0–180° and relaxing the rest of molecular parameters.
The resulting energy surface was fitted to a suitable expansion of
trigonometric functions into the coupled, two-top model. The
equations employed show a Fourier expansion like Eq. (1) for each

66
G. García et al. / Chemical Physics 374 (2010) 62–76
Fig. 2. Rotational Barriers for 2-hydroxystyrene at B3LYP/6-31G* (
D), mPW1PW91/6-31G* (h) and MP2/6-31G* (J). (a) From conformer 1 to conformer 4. (b) From
conformer 1 to conformer 3. (c) From conformer 3 to conformer 2. (d) From conformer 4 to conformer 2.
Fig. 3 shows the PES at B3LYP/6-31G* while Table 3S shows the Vi
rotor, and several other terms, which describe the perturbation be-
tween rotors. There are different expressions for these coupling
terms; in this paper we have used the terms suggested by Durig
*
parameters both for B3LYP and MP2 with 6-31G . The goodness of
the fitting procedure has been assessed through the correlation
*
coefficient, R2.
and Zheng [29]. The resulting equation for B3LYP(MP2)/6-31G
can be written as
2ð4Þ
4
X
X
1
2
1
2
3.4. Internal barrier decomposition schemes
Vð
a; bÞ ¼
Við1 ꢁ cosði
a
ÞÞ þ
Vjð1 ꢁ cosðjbÞÞ
i¼1
j¼1
In order to investigate the effects governing the torsion barriers
and preferred conformations, the vinyl and hydroxyl torsional bar-
riers have been featured using two different schemes. In the first
one, the total energy changes have been decomposed as a sum of
potential and kinetic contributions. In the second one, the natural
bond orbital, NBO, partitioning scheme has been applied in order
2ð3Þ
2ð3Þ
X
X
þ
þ
VIkIðcosðk
a
Þ cosðkbÞ ꢁ 1Þ þ
VItII sinðt
Þ sinðtbÞ
a
t¼1
k¼1
2ð3Þ
2ð3Þ
X
X
VIrV cosðr
a
Þ sinðrbÞ þ
VVs sinðs
aÞ cosðsbÞ
ð2Þ
r¼1
s¼1

G. García et al. / Chemical Physics 374 (2010) 62–76
67
Table 2
where
phenyl–hydroxyl and phenyl–vinyl dihedral angles,
change for the nuclear repulsion energy, Ene is the change for
the electron-nuclear attraction energy, Eee is the change for the
Ek is the change for the kinetic
D
E stands for the total energy change as a function of the
Rotational barrier height (kcal molꢁ1).
D
Enn is the
D
Torsional
Barrier
Method
Torsional
Barrier
Method
D
electron repulsion energy and
energy.
D
a
D
D
D
D
D
D
E1
E2
E3
E4
E5
E6
2.72
2.54
1.72
1.50
1.25
0.65
B3LYP
mPW1PW91
MP2
B3LYP
mPW1PW91
MP2
c
2.70
3.67
MP2
MP2
Table 3 collects the values for
DEee, DEnn and DEne for each bar-
rier as a function of the theory level while Fig. 4 shows the same as
a function of the C4–C3–C1–C2/H17–O16–C2–C5 dihedral angles
*
at MP2/6-31G level. The remaining levels of theory show a similar
behaviour.
b
D
D
D
D
D
D
E1
E2
E3
E4
E5
E6
4.59
4.97
5.27
3.94
3.91
3.32
B3LYP
mPW1PW91
MP2
B3LYP
mPW1PW91
MP2
d
3.81
3.91
MP2
MP2
The first, general conclusion from this figure is that the repul-
sive terms, i.e.
the dihedral angle varies, especially in the case of the vinyl dihe-
dral angle. The attractive term, Ene, for barriers a and c increases
DEnn and DEee evolve in a very similar fashion as
D
as the vinyl dihedral angle departs from planarity until its maxi-
mum value (note that the attractive term is negative) at ca. 40°
changing this trend rather monotonously to the minimum value
8
6
appeared at
to
ꢄ 40°, wherein they get their maximum values, and after-
wards they diminish getting negative values from
ꢄ 80° (barrier
a) and = 60° (barrier c). Thus, barriers a ( = 100°) and c ( = 80°)
a
ꢄ 180°. The repulsive terms, however, increase up
a
B3LYP/6-31G*
a
a
a
a
may be thought as mainly due to a dropping of the attractive term
rather than to an increase of the repulsive terms. Also, conformer 1
seems to be favoured by a steep increase of the attractive term
4
2
(
(
a
a
ꢄ 40°, barrier a) while conformers 2 (
a = 150°, barrier c) and 4
= 140°, barrier a) look to be stabilized by a diminution in the
repulsive terms. For conformer 3 both, attractive and repulsive
terms appear to be equally weighted.
The behaviour for barriers b and d is rather different since the
attractive term diminishes until b ꢄ 100° (barrier b) and 90° (bar-
rier d), increasing afterwards until b = 180°. The repulsive terms
follow the same pattern. Thus, in both cases, the main barrier-
forming factor looks to be the diminution in the attractive term,
whose minimum value appears at b ꢄ 100° (barrier b) and
b ꢄ 90° (barrier d). As a result the conformer 1 appears to be stabi-
lised by a slight diminution in the repulsive terms, conformer 2 is
favoured by an increasing in the attractive term as conformer 3
and, finally, conformer 4 looks to be stabilised by a slight diminu-
tion of both the attractive and repulsive terms.
0
-2
-4
0
20
40
60
80
100
120
140
160
180
20
40
60
80
100
120
140
160
180
A complementary analysis of the nature of the barriers can be
performed by analysing the Lewis and hyperconjugative (non-Le-
wis) terms according to Eq. (4) on the basis of the natural bond
orbital, NBO, method [8]
Fig. 3. PES for 2-hydroxystyrene at B3LYP/6-31G* level.
to decompose the total energy in the ELewis and Edeloc terms. These
partitioning schemes have been applied at MP2, B3LYP and
mPW1PW91 (these two only for barriers a and b) levels with 6-
31G* basis set.
According to the first approach, the energy of the torsional bar-
rier can be written as
D
E ¼
D
ELewis
þ
D
Edeloc
ð4Þ
where
D
ELewis stands for the energy of the hypothetical, localized
species described by a determinant of nearly doubly occupied NBOs
comprising the core, lone-pairs and localized bonds of the Lewis
D
E ¼
D
Enn
þ
D
Ene
þ
D
Eee
þ
D
Ek
ð3Þ
structure and
DEdeloc stands for the delocalisation energy change,
Table 3
D
Enn
,
D
Ene and D
Eee (kcal molꢁ1) for the different conformers and barriers.
*
*
*
MP2/6-31G
B3LYP/6-31G
mPW1PW91/6-31G
D
Enn
D
Ene
D
Eee
D
Enn
D
Ene
D
Eee
D
Enn
D
Ene
D
Eee
a
b
c
1
4
468.48
ꢁ888.36
11.30
ꢁ403.33
ꢁ978.63
507.43
ꢁ966.03
12.18
ꢁ253.09
536.85
ꢁ943.61
15.77
ꢁ480.81
ꢁ1124.80
584.97
ꢁ855.88
16.29
ꢁ333.04
560.80
ꢁ791.00
11.11
ꢁ547.59
ꢁ1185.51
534.37
ꢁ704.90
11.27
ꢁ412.18
1852.80
1797.98
1484.76
1
3
ꢁ24.50
ꢁ32.67
ꢁ21.67
656.00
814.93
969.87
3
2
ꢁ0.27
0.54
1754.91
ꢁ0.35
ꢁ877.87
ꢁ882.32
d
4
2
33.39
ꢁ51.28
ꢁ75.64
ꢁ0.01
41.21
55.69

68
G. García et al. / Chemical Physics 374 (2010) 62–76
Fig. 4. Dependence of
Eee (d) in kcal molꢁ1 as a function of vinyl or hydroxyl torsional angle at MP2/6-31G level. (a) From conformer 1 to
*
DETotal (–), DEnn (O), DEne (j), and D
conformer 4. (b) From conformer 1 to conformer 3. (c) From conformer 3 to conformer 2. (d) From conformer 4 to conformer 2.
i.e. the hyperconjugative stabilization arising from bond ? anti-
bond occupancy transfer.
such a manner that they almost counterbalance each other at
any value of phenyl–hydroxyl dihedral angle. Finally, barrier d also
shows the general barrier-forming and anti barrier-forming char-
Since the general appearance of both components,
Edeloc, as a function of the phenyl–vinyl and phenyl–hydroxyl
dihedral angles is quite similar for the different levels of theory,
DELewis and
D
acter of
DEdeloc and DELewis, respectively showing their extreme
values at b ꢄ 90°.
*
we will only detail in some extent the results at MP2/6-31G
(see Fig. 5). Table 4 shows the numerical values for E, ELewis E-
deloc for the different conformers for each method and each barrier.
Accordingly, in all the cases, Edeloc favours planar conformations
while ELewis favours those forms where both rotors are approxi-
The most important hyperconjugative interactions at MP2/6-
31G* level appear in Table 5 whereas the full results can be seen
D
D
, D
*
in Table 4S. We have selected MP2/6-31G because this is the only
D
method which predicts the four conformers with Pople’s basis sets.
The largest hyperconjugative interactions are those wherein the
benzene ring takes part. These interactions are similar for all the
conformers. As can be seen, the main NBO interactions between vi-
D
mately perpendicular to the benzene ring.
The energy profiles turn out to be, in general, rather barrier-
dependent. Thus,
up to = 90° (barrier a) and
wards in a monotonous trend up to
term, Edeloc, however, shows a continuous increase up to
to diminish next in a monotonous pattern up to = 180° (actually,
barrier a shows a shallow decrease at the very beginning of the
departure from planarity up to Edeloc is the
ꢄ 10°). As a result,
barrier-forming term for both barriers, i.e. a and c.
The phenyl–hydroxyl barriers look rather different. Thus, while
for barrier b Edeloc and
ELewis intercross each other at b ꢄ 135°,
this is not the case for barrier d whatever the phenyl–hydroxyl
dihedral angle. Also, Edeloc becomes lower than ELewis when
b < 20° and b > 135° while this is not the case for barrier d. As a
consequence, the marked barrier-forming character of Edeloc for
b < 135° changes into an anti barrier-forming character when
b > 135° for barrier b. The opposite trend is shown by ELewis in
D
ELewis shows a continuous descent from
= 100° (barrier c) increasing after-
= 180°. The hyperconjugative
= 90°
a
= 0°
nyl and benzene ring,
r
(C3–C4) ?
r (C–C)ring, are for the con-
*
a
a
formers 2 and 3 wherein the vinyl moiety deviates less from
planarity; the occupancy transfer between the hydroxyl and the
benzene ring, however, does not follow the same trend since the
most important are for conformers 1 and 2. These interactions
a
D
a
a
are LP(O16) ?
r
*(C–C)ring, favouring the electronic delocalization
a
D
in this region, and as a consequence, increasing the double-bond
character of the C–O bond. In addition, for conformer 3, a new
interaction involving the lone pair of the oxygen, LP(1)O16 and
the C4–H11 bond appears.
D
D
D
D
3.5. Topological analysis of the electronic density
D
One final attempt was made to rationalize the structure of 2-
hydroxystyrene using Bader’s Atoms-in-Molecules (AIM) theory.
D

G. García et al. / Chemical Physics 374 (2010) 62–76
69
*
DETotal (s), DELewis (j), and DEdeloc (p) with each rotor at MP2/6-31G level. (a) From conformer 1 to conformer 4. (b) From conformer 1 to conformer 3.
Fig. 5. Dependence of
(c) From conformer 3 to conformer 2. (d) From conformer 4 to conformer 2.
Table 4
DE, DELewis, D
Edeloc (kcal molꢁ1) for the different conformers and barriers.
*
*
*
MP2/6-31G
B3LYP/6-31G
mPW1PW91/6-31G
D
E
D
ELewis
D
Edeloc
D
E
D
ELewis
D
Edeloc
D
E
D
ELewis
D
Edeloc
a
1
4
ꢁ1.996
ꢁ0.77
ꢁ4.29
ꢁ7.74
2.29
6.97
ꢁ2.34
ꢁ1.04
ꢁ5.10
ꢁ8.59
2.76
7.55
ꢁ2.95
ꢁ2.42
ꢁ5.64
ꢁ8.24
2.69
5.81
b
c
1
3
ꢁ0.29
0.06
1.98
ꢁ0.24
ꢁ1.48
ꢁ0.22
ꢁ0.07
ꢁ0.16
ꢁ1.60
ꢁ0.16
0.19
6.26
ꢁ0.35
ꢁ5.43
0.07
7.74
0.51
0.91
2.51
0.84
3
2
0.00
ꢁ0.97
ꢁ0.07
ꢁ8.71
d
4
2
ꢁ0.32
ꢁ0.35
ꢁ0.69
ꢁ0.37
0.38
0.01
Table 5
There are four types of critical points, but only Bond Critical Points
(BCP) and Ring Critical Points (RCP) will be analysed in this work.
In particular, we will only focus our attention on those BCPs related
with the existence of a hydrogen bond. According to the AIM the-
Most significant MP2/6-31G NBO interactions (kcal molꢁ1).
*
1
2
3
4
r
(C3–C4) ?
r
*(C–C)ring
9.45
42.79
3.98
–
11.70
40.5
–
14.40
33.76
–
6.79
39.61
–
LP(O16) ?
r
*(C–C)ring
r
ory, BCPs are characterized through the charge density (
q
), its
Laplacian ( ), defined as
r
2q), and the ellipticity (
e
e
= |(k1/k2) ꢁ 1|,
r
(C3–C4) ?
*(O16–H17)
*
LP(O16) ?
r
(C4–H11)
–
2.21
–
where k1 and k2 are the largest and the smallest curvatures of
the charge density in a direction perpendicular to the bond path,
respectively. Ellipticity measures the extent to which charge is
preferentially accumulated. It provides a criterion for structural
stability in such a manner that the larger the ellipticity the weaker
the bond [30]. Also, in order to analyse RCPs we use the total en-
In this context, the localization of the critical points (CP) is a very
suitable tool for the characterization of the molecular electronic
structure in terms of the nature and magnitude of the interactions.

70
G. García et al. / Chemical Physics 374 (2010) 62–76
ergy density (H), as a quantitative measurement of
delocalisation [31].
A molecular graph (see Fig. 6) for each conformer was built at
MP2 and DFT (B3LYP and mPW1PW91) methods in combination
p
-electron
very weak intramolecular contact as the hydroxyl moiety dimin-
ishes the -electronic density of the vinyl group (acceptor group)
[28,33,34]. Baker and Shulgin [26] also studied this contact and
they noted that the low basicity of -electrons in 2-hydroxysty-
p
p
*
**
with 6-31G and 6-311++G basis sets, to locate the BCPs and
the RCPs, by means of the AIM2000 program. All conformers show,
as expected, a RCP (namely RCP 1) for the benzene ring. Table 5S
contains the values for the different AIM parameters for the differ-
ent conformers. As can be seen, the relative orientation of the vinyl
and hydroxyl moieties does not play any critical role in the elec-
tronic delocalisation within the benzene ring given the similarity
between the AIM parameters.
rene makes the interaction to be very weak.
After a comparison between BCP 3 and BCP 4 we conclude about
the similarity between their electronic densities while the elliptic-
ity is larger for conformer 3, pointing this interaction is stronger for
conformer 1. It must be stated that these conclusions are sup-
ported by NBO results since, as appeared in Table 5, interactions
*
*
(C4–H11) become
r
(C3–C4) ?
r
(O16–H17) and LP(O16) ?
r
3.98 and 2.21 kcal molꢁ1 for conformers 1 and 3, respectively,
while these interactions are absent in any other case.
According to the molecular graphs, conformers 1 and 3 show a
bond path between H17 and C4 (BCP 3) and O16-H11, (BCP 4),
respectively. Tables 6 and 7 collect AIM parameters for both BCPs.
The electronic density (
0.0152 to 0.0170 au and
an intramolecular interaction [32]. The result for conformer 1 sup-
ports the conclusion by Oki and Iwamura who found an intramo-
lecular hydrogen bond between the hydroxylic hydrogen (H17)
Conformers 1 and 3 also show a pseudo-ring with their corre-
sponding ring critical points, i.e. RCP 2 and RCP 5, respectively.
The size of the pseudo-ring will contribute to the formation of
the hydrogen bond. Assuming that for conformer 1 the interaction
q
r
) shows a value within the range from
2q > 0, that fulfils the criteria for such
occurs between H17 and (C3–C4) p-electrons, a pseudo-ring of six
members will result. Also, for conformer 3 the contact O16ꢀ ꢀ ꢀH11
renders a six-membered pseudo-ring coplanar to the benzene ring.
Table 8 shows the AIM parameters for these RCPs.
and the
p-electrons of the vinyl group, giving rise to an unusual,
Fig. 6. Molecular graphs for each conformer obtained with AIM2000 program.
Table 6
Local properties of BCP3 for conformer 1 (
q
, r2q, ki,
e in atomic units).
q
r
2q
k1
k2
k3
e
B3LYP
6-31G*
6-311++G
0.0159
0.0154
0.0596
0.0542
ꢁ0.0151
ꢁ0.0145
ꢁ0.0158
ꢁ0.0156
ꢁ0.0142
ꢁ0.0138
ꢁ0.0079
ꢁ0.0060
ꢁ0.0088
ꢁ0.0064
ꢁ0.0076
ꢁ0.0053
0.0826
0.0747
1.0146
1.0172
**
**
*
mPW1PW91
MP2
6-31G
0.0168
0.0165
0.0612
0.0574
0.0859
0.0793
1.0141
1.0184
6-311++G
6-31G*
6-311++G**
0.0161
0.0159
0.0573
0.0547
0.0791
0.0739
1.0133
1.0170

G. García et al. / Chemical Physics 374 (2010) 62–76
71
Table 7
Local properties of BCP4 for conformer 3 (
q
, r2q, ki,
e in atomic units).
q
r
2q
k1
k2
k3
e
*
B3LYP
6-31G
0.0158
0.0152
0.0601
0.0620
ꢁ0.0160
ꢁ0.0147
ꢁ0.0150
ꢁ0.0144
0.0911
0.0911
1.0019
1.0006
**
6-311++G
*
mPW1PW91
MP2
6-31G
0.0163
0.0157
0.0625
0.0647
ꢁ0.0166
ꢁ0.0154
ꢁ0.0171
ꢁ0.0156
ꢁ0.0156
ꢁ0.0149
ꢁ0.0162
ꢁ0.0138
0.0946
0.0950
1.0020
1.0009
6-311++G**
6-31G*
6-311++G**
0.0170
0.0156
0.0644
0.0640
0.0978
0.0935
1.0018
1.0035
Table 8
sis set whatever the method chosen so that the electron delocaliza-
tion will be enhanced in conformer 3. As shown in Fig. 5, the
delocalization energy (Edeloc) mainly favours conformer 3 rather
than conformer 1. Besides, since the electron delocalization is
Local properties of RCP 2 and RCP 5 (
q
, r2q, G and V in atomic units; i, in kcal molꢁ1).
q
r
2q
H
G
ꢁV
RCP 2
greater in RCP 5, the electron density, q, will be lower for this pseu-
do-ring as shown in Table 8.
B3LYP
6-31G*
0.0130
0.0667 2.0080 0.0135 0.0103
6-311++G** 0.0135 ꢁ0.0349 1.8198 0.0127 0.0098
*
mPW1PW91 6-31G
0.0141
0.0146
0.0719 2.0708 0.0147 0.0114
0.0680 1.8825 0.0140 0.0110
**
6-311++G
3.6. Vibrational analysis
MP2
6-31G*
6-311++G** 0.0150
0.0144
0.0697 1.7570 0.0146 0.0118
0.0668 1.6315 0.0141 0.0115
3.6.1. Scaled quantum mechanical force field and vibrational
assignment
RCP 5
B3LYP
6-31G*
6-311++G** 0.0099
0.0099
0.0559 1.7570 0.0112 0.0084
0.0582 1.9452 0.0114 0.0083
In order to perform a complete vibrational assignment of the
fundamental modes, the SQM methodology was employed and
then the infrared spectra of the different conformations of 2-
hydroxystyrene were generated as follows. First, we have scaled
mPW1PW91 6-31G*
6-311++G
0.0103
0.0102
0.0585 1.7965 0.0118 0.0089
0.0609 2.0708 0.0120 0.0087
**
**
*
*
MP2
6-31G
0.0105
0.0103
0.0588 1.6943 0.0121 0.0094
0.0602 1.8825 0.0120 0.0090
the appropriate B3LYP/6-31G quadratic force constant matrix in
6-311++G
a natural internal coordinates basis [21] using a set of transferable
scale factors [36]. The SQM methodology uses a full set of non-
redundant, natural coordinates that are sorted into groups which
share a common scaling factor. The initial set of theoretical force
constants FTheo is scaled by a diagonal matrix S which contains
Accordingly, the density of total energy H for RCP 5 becomes
**
slightly greater than that for RCP 2 when using the 6-311++G ba-
Fig. 7. Internal coordinates as defined for 2-hydroxyestyrene.

72
G. García et al. / Chemical Physics 374 (2010) 62–76
Table 9
Assignments of the fundamental modes and those scaled at B3LYP/6-31G* for 2-hydroxystyrene (cmꢁ1).
mexp (cmꢁ1
)
mSQM (multiple factors)
Approximate description
1
3
4
3610sIR
3602
(O–H) st (4)
3562wIR
3474mIR
3088wIR
3072mIR
3066mIR
3020mRa
2967mIR
2932mIR
3592
3153
(O–H) st (3)
(O–H) st (1)
(C4–H) st (3)
(C4–H) st (1, 4)
(C–H)b st (1, 3, 4)
(C–H)b st (1, 3, 4)
(C–H)b st (1)
(C4–H) st (1, 4)
3530
3114
3078
3061
3049
3039
3116
3079
3066
3082
3066
3030
3012
3039
2991
(C–H)b st (3)
(C3–H) st (1, 3)
(C3–H) st (4)
(C4–C3) st, (C4–H) bending, (C3–H) rock (1)
(C–C)b st, benzene asym. def. (1, 4)
(C–C)b st, (Cb–H) rock (1); (C–C)b st, (Cb–H) rock, benzene asym. def. (4)
(C–H)b rocking, (C–C)b st (3)
(C–H)b rocking, (C–C)b st (1); (C–H)b rocking, (C–C)b st, (C–H)b rocking (4)
(C–H)b rocking, (C–C)b st (1, 4)
(C4–H) bending, (C–H)b rocking (3)
(C4–H) bending, (C3–H) rocking, (C–H)b rocking (1, 3, 4)
(C–O–H) bending, (C–H)b rocking (1)
(C–O–H) bending, (C–C)b st, (C–H)b rocking (4)
(C–O–H) bending, (C–H)b rocking, (C–H)b rocking (1, 4)
(C–C)b st (1); (C–H)b rocking, (C–C)b st (3); (C–C)b st, (C–H)b wagging (4)
(C–O) st, Benzene Trig. Def., (C–C)b st (1); (C–O) st, (C–H)b rocking, (C-C)b st (4)
(C–H)b rocking, (C–O–H) bending, (C–C)b st (1)
(C–O–H) bending, (C–O) st, benzene trig. def., (C–C)b st (4)
(C–C)b st, (C–H)b rocking, (C–O–H) bending (1)
(C–O–H) bending, (C–C)b st (3); (C–H)b rocking, (C–O–H) bending, (C–C)b st (4)
(C–H)b rocking (1, 4)
2901mIR
2876mIR
1628mIR
1606mIR
1582mIR
1502mIR
1485mIR
1455vsIR
3018
1632
1617
1577
1617
1580
1505
1464
1412
1488
1467
1485
1464
1422wIR
1350mIR
1329wRa
1312wIR
1295mIR
1240mIR
1222shIR
1214sRa
1180sIR
1413
1352
1414
1342
1309
1296
1250
1315
1294
1245
1224
1296
1169
1205
1170
1170shIR
1157mIR
1117wIR
1096mIR
1054wIR
1031vwIR
1021wIR
1169
1152
1155
1109
(C–C)b st, (C–H)b rocking (1)
1113
1040
(C–C)b st, (C–H)b rocking, (C–O–H) bending (3)
(C–C)b st (3); (C4–H) rocking, (C3–H) rocking, (C–C)b st (4)
(C4–H) rocking, benzene trig. def. (1)
1044
1016
1030
1027
(C–C)b st, (C–H)b rocking (1);
(C4–H) rocking, Vinyl rocking (C–C)b st (4)
(C3–C4) wagging, (C3–C4) tors. (1, 3, 4)
(C–H)b wagging, benzene puck (1, 4)
(C–H)b wagging, asym. tors. benzene (1, 4)
Vinyl wagging (1, 3, 4)
(C–H)b wagging, benzene puck (1); (C–H)b wagging (4)
Benzene trig. def., (C–O) st (1, 4); (C–H)b wagging, benzene puck (3)
(C–H)b wagging (1, 4)
Benzene puck., (C–O) wagging, (C–H)b wagging (1, 3); benzene puck (C–O) wagging, benzene wagging (4)
Benzene puck., (C–O) wagging (4)
(C–C)b st, benzene puck., sym. def. benzene (1); (C–C)b st, sym. def. benzene (3)
Benzene Puck, (C3–C4) tors., (C–O) wagging (1)
Sym. def. benzene, vinyl sym. def. (C–O) rocking (1); sym. def. benzene (4)
Sym. Def. Benzene (1); benzene sym. def., vinyl rocking, (C–C)b st (3)
Asym. tors. benzene, benzene puck (C–O) wagging (4)
Asym. Tors. Benzene, (C–O) wagging, Benzene Puck. (1, 3)
Sym. def. benzene, vinyl sym. def. (C–O) wagging (4)
(C–O) rocking (4)
998sIR
990
968
935
923
865
846
766
747
989
983
967
937
924
864
839
765
746
731
971vwIR
936shIR
915mIR
861shRa
840mIR
762sIR
925
841
745
718
751sIR
728vwIR
717vwIR
645Ra
720
651
597
566
589wIR
568wRa
539vwIR
530wRa
509wIR
508wIR
434vwRa
426wIR
409wIR
390shRa
382shRa
368shRa
311shRa
265shRa
255shRa
233mRa
220wRa
157mRa
135shRa
104sRa
593
536
568
530
529
509
502
439
409
384
Asym. tors. benzene, benzene wagging (1)
Asym. tors. benzene, benzene wagging (3, 4)
(C–O) rocking, sym. def. benzene (1)
429
390
353
424
Vinyl sym. def., (C–O) rocking (3)
Asym. tors. benzene, vinyl sym. def. (C–O) rocking (1)
Hydroxyl tors. (3); Hydroxyl tors., (C–O) rocking, sym. def. benzene (4)
Hydroxyl tors. (4)
Asym. tors. benzene, benzene wagging, vinyl rocking (4)
Asym. tors. benzene, benzene wagging (1, 3)
Vinyl rocking, vinyl sym. def. (1)
Asym. tors. benzene, vinyl rocking, vinyl sym. def. (4)
Vinyl tors., asym. tors. benzene (1)
Asym. tors. benzene, vinyl rocking, benzene wagging (4)
Vinyl tors., hydroxyl tors., vinyl sym. def. (1)
Vinyl tors., (C3–C4) wagging (3); vinyl tors. (4)
375
316
265
259
235
258
226
142
88
159
107
84sRa
83
s = strong, m = medium, w = weak, sh = shoulder, v = very.
the scaling factors according to the matrix relationship FScal = S1/2
FTheo 1/2, in such a manner that any diagonal force constant Fii is
scaled by the scaling factor Sii while the non-diagonal force con-
stants Fij is scaled by (SiiSjj)1/2. Thus we used the set of 12 scale fac-
S

G. García et al. / Chemical Physics 374 (2010) 62–76
73
tors proposed by Rauhut and Pulay [36], optimized at B3LYP/6-
31G* level (multiple scaling). In addition, we have also applied
the so called uniform scaling, that is the scaling with a unique,
optimized scale factor (0.928) that was also proposed by Rauhut
and Pulay [36]. In order to simulate the IR spectrum, the solution
of the El‘yashevich [37]–Wilson [38] vibrational secular equation
resulted the modified set of the fundamental frequencies and IR
intensities [39]. We have used the SCALE3 program [23] for scaling
of the force fields and for the solution of the vibrational secular
equation. Finally, Lorentzian band shapes [40] were supposed with
the same value of half-width (12 cmꢁ1) for all the bands [39].
The vibrational assignment of 2-hydroxystyrene was carried
out by combining the experimental IR (neat liquid and Cl4C solu-
tion) and Raman (neat liquid) spectra with those obtained from
the SQM procedure outlined above. No correction on the optimised
geometry has been applied. Fig. 7 shows the internal coordinates
for 2-hydroxystyrene and Table 6S collects the natural coordinates
and their scaling factors. Theoretical as well as the uniformly and
multiply scaled frequencies for conformers 1, 3 and 4 are also col-
lected as Supplementary material (Table 7S). Table 9 collects all the
experimental and SQM frequencies (obtained by multiple scaling)
for the likely conformers as well as an approximate description
of the vibrational modes attained at through the potential energy
distribution (PED) [41]. Only modes with a contribution larger than
10% have been taken into account.
Fig. 4S shows the IR spectrum for the pure liquid and CCl4 solu-
tion and Fig. 5S(a) and (b) shows the Raman spectra for the pure
liquid of 2-hydroxystyrene. In a previous work, Fateley et al. [3]
determined the number of existing conformers from the number
of O–H stretching bands. Furthermore, the width of IR bands in
a
400
500
600
700
800
900
1000
1100
1000
1100
1200
1300
1400
1500
1600
1700
Wavenumber (cm-1)
Fig. 8. (a). The infrared spectra of the conformers 1 (
),3 (ꢀ ꢀ ꢀ) and 4 (—). (a.1) Experimental spectra. Region 400–1100 cmꢁ1. (a.2) SQM spectra using multiple scaling at
B3LYP/6-31G level. Region 400–1100 cmꢁ1. (a.3) Experimental spectra. Region 1000–1700 cmꢁ1. (a.4) SQM spectra using multiple scaling at B3LYP/6-31G level. Region
*
*
1000–1700 cmꢁ1. (b) The infrared spectra of the conformers 1 (
), 3 (ꢀ ꢀ ꢀ) and 4 (—). (b.1) Experimental spectra. Region 2800–3800 cmꢁ1. (b.2)ꢁS1QM spectra using multiple
ꢁ1
*
*
scaling at B3LYP/6-31G level. Region 2950–3175 cm . (b.3) SQM spectra using multiple scaling at B3LYP/6-31G level. Region 3490–3620 cm
.

74
G. García et al. / Chemical Physics 374 (2010) 62–76
b
2850
3000
3150
3300
3450
3600
3750
2960
3000
3040
3080
3120
3160
3500
3520
3540
3560
3580
3600
3620
Wavenumber (cm-1)
Fig. 8 (continued)
the high frequency region may be thought as due to the existence
of hydrogen bond contacts [35].
IR spectra in CCl4 solution, the authors found similar shifting for
the O–H stretching mode. Thus, for 2- and 4-methoxyphenol the
O–H stretching mode is displaced from 3618 (OH free) to 3558 (in-
tra) cmꢁ1, respectively. In the case of 2- and 4-nitrophenols [42], a
After a careful analysis of the O–H stretching region in the
infrared spectrum in CCl4 solution, we could clearly assign more
than one conformer. As can be seen in Fig. 8(b.1), three bands
clearly resolved appear at 3474, 3562 and 3610 cmꢁ1. After apply-
ing the SQM methodology, two bands were calculated at 3530 and
3592 cmꢁ1 which could be tentatively assigned to the O–H stretch-
ing mode according to the corresponding PED matrix for conform-
ers 1 and 3, respectively. The intramolecular hydrogen bond in
conformer 1 produces an O–H stretching band which exhibits a
rather broad band with a maximum at 3474 cmꢁ1 (Fig. 8(b.1)). This
mode occurs at a lower frequency than that for the rest ꢁo1f conform-
dramatic shift was detected from 3596 (free) to 3243 (intra) cmꢁ1
.
As it can be seen in Fig. 8(b.2), the assignment of the C–H
stretching is more complicated in this region due to the overlap-
ping of the bands for the tree conformers. Furthermore, the band
observed at 1096 cmꢁ1 in the infrared spectrum could be assigned
to conformer 3 from the comparison between the experimental
and theoretical infrared spectra. As it can be seen in Fig. 8(a.3)
and (a.4), a band with similar intensity is calculated for conformer
3 at 1113 cmꢁ1. In addition, the doublet band observed at 1485 and
1502 cmꢁ1 could be assigned to conformers 1–4 and 3, respec-
tively. The corresponding scaled values are 1488–1485 and
1505 cmꢁ1, respectively. Finally, the contribution of the modes cal-
culated at 1467, 1456 and 1464 cmꢁ1 of the conformers 1, 3 and 4
yields the intense band observed at 1455 cmꢁ1 in the infrared
spectrum.
In the low energy region where the OH torsion appears, some
bands can be assigned to this vibration. Thus, the band observed
at 104 cmꢁ1 (Raman) can be assigned to the OH torsion in con-
former 1 while the bands observed at 368 and 382 cmꢁ1 in Raman
can be assigned to the OH torsion in conformers 3 and 4, respec-
tively. The bands observed at 104 and 84 cmꢁ1 in Raman can be
mainly assigned to the vinyl torsion movement for the conformers
1 and 3–4, respectively. As stated before, for conformer 1 there is
*
ers predicted by B3LYP/6-31G . The band at 3562 cm could be
assigned to conformer 3. In this conformer, a weak CH = CHꢀ ꢀ ꢀOH
contact is formed which could move the O–H stretching to a lower
wavenumber than that of the completely free O–H group.
The band at 3610 cmꢁ1 is fairly narrow compared with the
bands of other phenol derivatives possessing intra- and inter
molecular hydrogen bonds. In the case of conformers 2 (not pre-
dicted as a stationary point at B3LYP/6-31G* level) and 4, for which
a band at 3602 cmꢁ1 is predicted after scaling, the O–H group is
free. Thus, the shift of the O–H stretching band from 3610 (OH free)
to 3474 (intra) cmꢁ1 could be a consequence of the OHꢀ ꢀ ꢀ
p(vinyl
C@C) hydrogen bond formation. In a recent study about intra and
intermolecular hydrogen bond formation in some ortho-substi-
tuted phenols from an experimental and theoretical study of the

G. García et al. / Chemical Physics 374 (2010) 62–76
75
also a noticeable contribution of the hydroxyl torsion movement
that could point to the correlation between the hydroxyl and vinyl
torsions caused by the hydrogen bond between H17 (linked to
a prominent band is calculated for this one at 3699 cmꢁ1. Finally,
the conformer 4 is expected to be hidden under the contour of
the bands assigned to conformer 2 and 3. Thus, the experimental
IR spectral profile in the high energy region shows the existence
of, at least, three stable conformers, namely 1, 2 and 3. This result
is partially at odds with that from Berdyshev et al. [5] who ob-
tained a reversed intensity ratio for conformers 1 and 2 on the ba-
sis of Gibb’s free energy.
O16) and the
p electrons of the vinyl moiety.
3.6.2. Linear scaling
In order to obtain some additional information about the pres-
ence of other rotamers in the bulk conformational equilibrium of
2-hydroxystyrene, its vibrational spectrum has been monitored
through the appearance of distinctive IR spectral features in the
high frequency range. With this aim, we have calculated the IR
spectrum at B3LYP/cc-pVTZ level for which the full set of conform-
ers is expected and whose frequency scaling factor is known (see
Ref. [24]), as well. After scaling and normalizing the corresponding
intensities respect to conformer 1 taking into account the frac-
tional populations appearing in Table 1, the IR spectrum in the re-
gion 3500–4000 cmꢁ1 looks like Fig. 9(b).
4. Conclusions
The molecular structure, conformational preferences and vibra-
tional spectra has been investigated using IR/Raman techniques
and theoretical methods. MP2 method always predicts the
presence of four conformers while DFT (B3LYP and mPW1PW91)
methods only predict the presence of the four conformers when
Dunning’s correlation consistent basis sets are used. These
Accordingly, the observed bands at 3474 and 3562 cmꢁ1 may be
assigned again to the O–H stretching of conformer 1 and 3, whose
scaled wavenumbers are 3625 and 3689 cmꢁ1, respectively. Now,
the peak appearing at 3610 cmꢁ1 (OH free) could be attributed to
the conformer 2 according to the profile shown in Fig. 9(b) where
conformers are built by rotation around a simple
a consequence they are interchangeable. Conformer 1 (
r
bond and as
ꢄ 35°,
a
b ꢄ 10°) turns out to be the most stable regardless of the level of
theory with no correction of zero-point energy in mind. This result
partially matches that obtained by Berdyshev et al.’s [5] who ex-
tended their calculations to the domain of enantiomer couples as
well as to the effect of the thermal and entropy terms on the nature
of the global minimum. Thus, three pairs of stable, non-planar
enantiomers in addition to a flat rotamer result. Besides, as a con-
sequence of the thermal and entropy terms the nature of the global
minimum is now located on conformer 2. This effect may be ex-
plained on the basis of the small energy difference between the dif-
ferent extant conformers, most of which appear to be even smaller
a
than 1 kcal molꢁ1
.
Coupling between rotors has been studied through a two-
dimensional PES fitted to a suitable expansion of trigonometric
functions within the coupled, two-tops model.
The vinyl–phenyl and the hydroxyl–phenyl torsion barriers
have been analysed using MP2, B3LYP and mPW1PW91 method
*
with 6-31G basis set, in order to get some insight into the effects
governing the torsion barriers and the preferred conformation on
the basis of attractive/repulsive potential energy and hyperconju-
gative terms. According to the first scheme, conformer 1 seems
to be favoured by an increase of the attractive term, conformers
2 and 4 appear to be stabilized by a diminution in the repulsive
terms and for conformer 3, attractive and repulsive terms look to
3300
3400
3500
3600
3700
Wavenumber (cm-1)
be equally weighted. As to the second one,
forming term for barriers a and c, while for barrier b and d,
and ELewis almost counterbalance each other at any value of phe-
D
Edeloc is the barrier-
b
D
Edeloc
D
nyl–hydroxyl dihedral angle. The most important hyperconjuga-
tive interactions are those wherein the benzene ring takes part.
The magnitude of the hyperconjugative interactions is the highest
*
r C–C(ring).
for the LP(O16) ?
Hydrogen bonding has been found for the conformer 1 between
H17 and vinyl electron and between H11 and O16 for conformer
p
3 by using AIM theory. No conclusive result has been obtained
about the relative strength of such hydrogen bonds.
Infrared and Raman spectra have been recorded and analysed
for the neat liquid, and also CCl4 solution in the case of IR. Vibra-
tional assignment has been made on the basis of SQM methodol-
ogy. A careful analysis of the infrared spectrum compared with
the calculated at B3LYP/6-31G* level led us to conclude about the
presence of conformers 1, 3 and 4. On the other hand, after linear
scaling and intensity normalization, B3LYP/cc-pVTZ results predict
conformer 1 as the most abundant on the basis of the electronic
energy while Berdyshev et al. [5] predict the same for conformer
2 on the basis of Gibb’s free energy. As a conclusion, the vibrational
analysis shed light about the presence of at least three conformers.
3750
3550
3600
3650
3700
Wavenumber (cm-1)
Fig. 9. (a) The observed high frequency (3500–4000 cmꢁ1
)
IR spectrum of 2-
hydroxystyrene. (b) The B3LYP/cc-pVTZ high frequency IR spectrum of 2-hydroxy-
styrene after scaling and normalization. 1 (red), 2 (green), 3 (blue) and 4 (black).
(For interpretation of the references to colour in this figure legend, the reader is
referred to the web version of this article.).

76
G. García et al. / Chemical Physics 374 (2010) 62–76
[11] C. Møller, M.S. Plesset, Phys. Rev. 46 (1934) 618.
Acknowledgements
[12] J. Hehre, L. Radom, P.v.R. Schleyer, J.A. Pople, Ab Initio Molecular Orbital
Theory, Wiley, New York, 1986.
The authors gratefully acknowledge Consejería de Innovación,
Ciencia y Empresa, Junta de Andalucía (PAI-FQM 337 Contract
and FQM-P06-01864 Project) for financial support. The authors
also express their gratitude to Prof. Roser Pleixats, Departamento
de Química, Universidad Autónoma de Barcelona (Spain), for pro-
viding us with a sample of the title compound. One of us (G.G.)
thanks the Unidad Asociada CSIC-Universidad de Jaén for a pre-
doctoral grant. This work has been supported in part by the Hun-
garian Scientific Research Fund, OTKA, through Grant K68427.
[13] D.E. Woon, T.H. Dunning Jr., J. Chem. Phys. 98 (1993) 1358.
[14] R.A. Kendall, T.H. Dunning Jr., R.J. Harrison, J. Chem. Phys. 96 (1992) 6796.
[15] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
[16] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B 37 (1988) 785.
[17] J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 77 (1996) 3865 (Erratum:
Phys. Rev. Lett. 78 (1997) 1396).
[18] C. Adamo, V. Barone, J. Chem. Phys. 108 (1998) 664.
[19] E.D. Gledening, A.D. Reed, J.E. Carpenter, F. Weinhold, NBO, version 3.1.
[20] F. Biegler-Köning, J. Schönbohm, D. Bayles, J. Comput. Chem. 22 (2001) 545.
[21] G. Fogarasi, X. Zhou, P.W. Taylor, P. Pulay, J. Am. Chem. Soc. 114 (1992) 8191.
[22] P. Pulay, G. Fogarasi, G. Pongor, J.E. Boggs, A. Vargha, J. Am. Chem. Soc. 105
(1983) 7037.
[23] (a) G. Pongor, G. Fogarasi, I. Magdó, J.E. Boggs, G. Keresztury, I.S. Ignatyev,
Spectrochim. Acta A 48 (1992) 111;
Appendix A. Supplementary material
(b) Program SCALE3, Department of Theoretical Chemistry, Eötvös Loránd
University, Budapest, 1992.
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.chemphys.2010.06.014.
[24] J.P. Merrick, D. Moran, L. Radom, J. Phys. Chem. A 111 (2007) 11683.
[25] E. Marsault, H.R. Hoveyda, M.L. Peterson, C. Saint-Louis, A. Landry, M. Vezina, L.
Ouellet, Z. Wang, M. Ramaseshan, S. Beaubien, K. Benakli, S. Beauchemin, R.
Deziel, T. Peeters, G.L. Fraser, J. Med. Chem. 49 (2006) 7190.
[26] A.W. Baker, A.T. Shulgin, Dow Chemical Company, Werstern Division 80
(1958) 5358.
References
[27] M.E. Tuttolomondo, A. Navarro, T. Peña, M.P. Fernández-Liencres, J.M.
Granadino-Roldán, S.F. Parker, M. Fernández-Gómez, Chem. Phys. 361 (2009)
94.
[1] A. Simperler, H. Lampert, W. Mikenda, J. Mol. Struct. 448 (1998) 191.
[2] L. Pauling, J. Am. Chem. Soc. 58 (1936) 94.
[3] G.L. Carlson, W.G. Fateley, J. Phys. Chem. 77 (1973) 1157.
[28] M. Oki, H. Iwamura, Bull. Chem. Soc. Japan 33 (1960) 681.
[29] J.R. Durig, C. Zheng, J. Mol. Struct. 690 (2004) 31.
[30] A.A. Ceacero-Vega, T. Peña-Ruiz, M. Fernández-Gómez, J.M. Granadino-Roldán,
A. Navarro, M.P. Fernández-Liencres, U.A. Jayasooriya, J. Phys. Chem. A. 111
(2007) 6406.
[31] M. Palusiak, T.M. Krygowski, Chem. Eur. J. 13 (2007) 7996.
[32] U. Koch, P.L.A. Popelier, J. Phys. Chem. 99 (1995) 9747.
[33] M. Oki, H. Iwamura, Bull. Chem. Soc. Japan 39 (1966) 470.
[34] M. Oki, H. Iwamura, T. Onoda, M. Nishida, Bull. Chem. Soc. Japan 39 (1966)
813.
[35] G.A. Jeffrey, An Introduction to Hydrogen Bonding, Oxford University Press,
Oxford, 1997.
[36] G. Rauhut, P. Pulay, J. Phys. Chem. 99 (1995) 3093.
[37] M.A. El‘yashevich, Dokl. Akad. Nauk. SSSR 28 (1940) 605 (in Russian).
[38] E.B. Wilson Jr., J. Chem. Phys. 9 (1941) 76.
[4] F. Dubnikova, A. Lifshitz, J. Phys. Chem. A 106 (2002) 9278.
[5] D.V. Berdyshev, V.P. Glazunov, V.L. Novikov, Russ. Chem. Bull. Int. Ed. 57 (2008)
510.
[6] N. Sadlej-Sosnowska, J. Phys. Chem. A 107 (2003) 8671.
[7] A.E. Reed, L.A. Curtiss, F. Weinhold, Chem. Rev. 88 (1998) 899.
[8] R.F.W. Bader, Atoms in Molecules, A Quantum Theory, Clarendon Press, Oxford,
1990.
[9] GaussView Release 3.0, Gaussian Inc., Pittsburg, PA, USA, 2000.
[10] Gaussian 03, Revision C.01, Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria,
G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, Jr., J. A.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci,
B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.;
Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda,
Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.;
Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.;
Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.;
Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.;
Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford,
S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.
.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara,
A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.;
Gonzalez, C.; Pople, J. A.; Gaussian, Inc., Wallingford CT, 2004.
[39] Zs. Kolos, D. Knausz, J. Rohonczy, E. Vass, Gy. Tarczay, G. Pongor, Chem. Phys.
318 (2005) 191.
[40] I.N. Levine, Molecular Spectroscopy, Wiley, New York, 1975.
[41] Y. Morino, K. Kuchitsu, J. Chem. Phys. 20 (1952) 1809.
[42] G. Litwinienko, G.A. DiLabio, P. Mulder, H. Korth, K.U. Ingold, J. Phys. Chem. A
113 (2009) 6275.