Conformational analysis of a trihydroxylated derivative of cinnamic acid - A combined Raman spectroscopy and Ab initio study

Article abstract of DOI:10.1016/j.molstruc.2004.02.019

A conformational analysis of 3-(3,4,5-trihydroxyphenyl)-2-propenoic acid (3,4,5-trihydroxycinnamic acid, THPPE), a trihydroxylated cinnamic acid analogous to caffeic acid (a natural compound often present in diet), was carried out by Raman spectroscopy co

Full text of DOI:10.1016/j.molstruc.2004.02.019

Journal of Molecular Structure 693 (2004) 103–118
www.elsevier.com/locate/molstruc
Conformational analysis of a trihydroxylated derivative of cinnamic
acid—a combined Raman spectroscopy and Ab initio study
S.M. Fiuzaa, E. Van Besiena,b, N. Milhazesc,d, F. Borgesc, M.P.M. Marquesa,e,
*
a
´ ´ ˆ
Unidade I and D ‘Quımica-Fısica Molecular’, Faculdade de Ciencias e Tecnologia, Universidade de Coimbra, 3000 Coimbra, Portugal
bDept. Chemie, Katholieke Univ Leuven, 3001 Heverlee, Belgium
c
´ ˆ ´
Quımica Organica, Fac. Farmacia, Univ Porto, 4050-047 Porto, Portugal
d
ˆ ´
Institute Sup. Ciencias Saude-Norte, 4580 Paredes, Portugal
e
´ ˆ
Dept. Bioquımica, Faculdade de Ciencias e Tecnologia, Univ. Coimbra, Ap. 3126, 3001-401 Coimbra, Portugal
Received 8 January 2004; revised 2 February 2004; accepted 2 February 2004
Abstract
A conformational analysis of 3-(3,4,5-trihydroxyphenyl)-2-propenoic acid (3,4,5-trihydroxycinnamic acid, THPPE), a trihydroxylated
cinnamic acid analogous to caffeic acid (a natural compound often present in diet), was carried out by Raman spectroscopy coupled to Ab
initio MO calculations. Apart from the optimised geometrical parameters for the most stable conformers of this compound, and for one of its
dimeric species, the corresponding harmonic vibrational frequencies, as well as potential-energy profiles for rotation around several bonds
within the molecule, were obtained. Twenty one distinct conformers were found for THPPE, the lowest energy ones—THPPE 1 and THPPE
2—displaying a completely planar geometry. The conformational preferences of this system were thus found to be mainly ruled by the
stabilising effect of p-electron delocalisation. At the light of these results, a complete assignment of the corresponding solid state Raman
spectra was performed.
q 2004 Elsevier B.V. All rights reserved.
Keywords: Hydroxycinnamic acid derivatives; Raman spectroscopy; Ab initio calculations; Conformational analysis
1. Introduction
compounds, which, in turn, was found to be strongly
dependent on their structural characteristics [2,18,19].
The knowledge of the conformational preferences of this
type of compounds is thus of the utmost importance as a
starting point for future studies aiming at the understanding
of the structure–activity relationships underlying their
biological activity, which may contribute to the wider
goal of developing new and more effective therapeutic
agents (e.g. anticancer drugs). Nevertheless, the reported Ab
initio molecular orbital calculations on phenolic acid
derivatives are very scarce [20–25], semiempirical quan-
tum-chemical methods being commonly used instead.
Moreover, even these studies have aimed exclusively at
the explanation of the structural dependency of the
antioxidant activity of these phenols (e.g. caffeic acid
[24]), focusing only on their most stable geometries.
Phenolic acid derivatives constitute a group of natural
compounds present in human diet in significative amounts,
which have long been known to display both antioxidant
(through their radical scavenging activity) and prooxidant
properties [1–3]. They are involved in numerous metabolic
reactions and are naturally occurring in many plant-derived
food products, were they are largely responsible for the
browning process [4–6]. Apart from being widely used as
antioxidant food additives [7,8], some of them were lately
found to behave as inhibitors of deleterious oxidative
processes—e.g. in the prevention of cardiovascular diseases
and inflammatory processes [9,10], or even in cancer
[11–17]. Consequently, in the last few years there has
been a growing interest in the understanding of the
mechanisms associated to the biochemical role of phenolic
In the present study, a complete conformational analysis
of 3-(3,4,5-trihydroxyphenyl)-2-propenoic acid (3,4,5-tri-
hydroxycinnamic acid, THPPE) 2 which is the trihydroxy-
lated analogue to trans-caffeic acid 2 was carried out, by
*
Corresponding author. Tel.: þ351-239-853-600; fax: þ351-239-853-
607.
E-mail address: pmc@ci.uc.pt (M.P.M. Marques).
0022-2860/$ - see front matter q 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.molstruc.2004.02.019

104
S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
both Raman spectroscopy and Ab initio MO calculations.
Both the corresponding ethyl ester 2 ethyl 3-(3,4,5-trihy-
droxyphenyl)-2-propenoate (ethyl 3,4,5-trihydroxycinna-
mate, ETHPPE) 2 and trans-caffeic acid were studied for
comparison purposes (namely aiming at a thorough spectral
assignment).
intensity of the most important fragments). Melting points
¨
were obtained on a Kofler microscope (Reichert Thermo-
var) and are uncorrected.
Other conditions. Thin-layer chromatography (TLC)
was carried out on precoated silica gel 60 F254, with a
layer thickness of 0.2 mm. The following systems were
used for analytical control: silica gel, ethyl ether and
chloroform/acetone (6:4). The spots were visualised
under UV detection (254 and 366 nm), iodine vapour
and FeCl3 (5%).
2. Experimental
2.1. Synthesis
2.2. Ab initio MO calculations
3-(3,4,5-trihydroxyphenyl)-2-propenoic acid (3,4,5-trihy-
droxycinnamic acid, THPPE) and ethyl 3-(3,4,5-trihydrox-
yphenyl)-2-propenoate (ethyl 3,4,5-trihydroxycinnamate,
ETHPPE). A suspension of 3,4,5-trihydroxybenzaldehyde
(1.0 g), malonic acid (1.0 g) or mono-ethylmalonate (1.2 g),
anhydrous pyridine (5 ml) and four drops of aniline was
stirred overnight, at 50 8C. After cooling, ethyl ether was
added (50 ml). The organic phase was washed with HCl-2N
and water, and dried over Na2SO4. The solvent was then
evaporated and theresidue was recrystallized from ethyl ether
to yield the pretended compounds as light yellow solids.
3,4,5-Trihydroxycinnamic acid (THPPE). Yield 35%;
FTIR ymax: (cm21): 3510, 3382, 1666, 1618, 1597, 1539,
1454, 1425, 1388, 1317, 1221, 1151, 1028, 980, 914, 831,
800, 721, 634, 596, 542, 484. 1H-NMR d: 6.09 (1H, d, J ¼
15:8; H(a)),6.57 (2H, s, H(2), H(6)), 7.32 (1H, d, J ¼ 15:8;
H(b)), 8.74 (1H, s, OH), 9.17 (2H, s, OH); 13C-NMR d:
107.5 C(2, 6), 115.2C(a), 124.6 C(1), 136.3 C(4), 145.1
C(b), 146.2 C(3, 5), 168.0 (CyO); EI-MS m/z (%): 196
(Mþz, 100), 179 (22); 152 (57), 133 (27), 105 (16), 78 (51),
63 (57); mp 186–188 8C.
Ethyl 3,4,5-trihydroxycinnamate (ETHPPE). Yield 65%;
FTIR ymax: (cm21): 3394, 1656, 1601, 1508, 1467, 1371,
1317, 1277, 1203, 1142, 1034, 976, 845, 634, 596, 594. 1H-
NMR d: 1.23 (3H, s, CH3),.4.15 (2H, q, CH2), 6.18 (1H, d,
J ¼ 15:8; H(a)),6.60 (2H, s, H(2), H(6)), 7.38 (1H, d, J ¼
15:8; H(b)), 8.80 (1H, s, OH), 9.18 (2H, s, OH); 13C-NMR
d: 14.4 (CH3), 59.8 (CH2), 107.7 C(2, 6), 114.2 C(a), 124.5
C(1), 136.5 C(4), 145.5 C(b), 146.2 C(3, 5), 166.6 (CyO);
EI-MS m/z (%): 224 (Mþz, 100), 196 (17); 179 (86), 152
(48), 133 (40), 105 (21), 77 (32); mp 176–179 8C.
Apparatus. Infrared spectra were recorded on an ATI
Mattson Genesis Series FTIR spectrophotometer using
potassium bromide disks; only the most significant absorp-
tion bands are reported ðymax:; cm21). 1H and 13C NMR data
were acquired, at room temperature, on a Bru¨ker AMX 300
spectrometer, operating at 300.13 and 75.47 MHz, respect-
ively. Dimethylsulfoxide-d6 was used as a solvent; chemical
shifts are expressed in d (ppm), relative to tetramethylsilane
(TMS) which was used as an internal reference; coupling
constants (J) are given in Hz. Assignments were also made
from DEPT (underlined values). Electron impact mass
spectra (EI-MS) were carried out on a VG AutoSpec
instrument; the data are reported as m/z (% of relative
The Ab initio calculations 2 full geometry optimisation
and calculation of the harmonic vibrational frequencies 2
were performed using the GAUSSIAN 98W program [26],
within the Density Functional Theory (DFT) approach, in
order to properly account for the electron correlation effects
(particularly important in this kind of conjugated systems).
The widely employed hybrid method denoted by B3LYP
[27–32], which includes a mixture of HF and DFT
exchange terms and the gradient-corrected correlation
functional of Lee, Yang and Parr [33,34], as proposed and
parametrised by Becke [35,36], was used, along with the
double-zeta split valence basis sets 6-31G* and 6-31G**
[37,38].
Molecular geometries (of both monomeric and dimeric
species) were fully optimised by the Berny algorithm, using
redundant internal coordinates [39]: The bond lengths to
within ca. 0.1 ppm and the bond angles to within ca. 0.18.
The final root-mean-square (rms) gradients were always less
than 3 £ 1024 hartree.bohr21 or hartree.radian21. No
geometrical constraints were imposed on the molecules
under study. The 6-31G** basis set was used for all
geometry optimisations, while for most of the rotational
energy barrier calculations the smaller basis 6-31G* was
found to yield good results. All frequency calculations were
run at the B3LYP/6-31G** level and wavenumbers above
400 cm21 were scaled [40] before comparing them with the
experimental data. The basis set superposition error (BSSE)
correction for the dimerisation energies was estimated by
counterpoise (CP) calculations.
Quantitative potential-energy profiles for rotation around
different bonds within the molecule were obtained, by
scanning the corresponding dihedrals and using least-
squares fitted Fourier-type functions of a dihedral angle, t
[41,42]:
3
2
X
X
1
2
1
2
P ¼
Pn½1 2 cosðntފ þ
P0msinðmtÞ
ð1Þ
n¼1
m¼1
The parameters Pn and P0m correspond to potential-energy
(Vn and Vm terms), bond-distance or bond-angle differences
relative to a reference value. According to the symmetry of
the molecule under study, the sine terms, which are of

S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
105
significance only for asymmetric functions around 1808,
were not included in the fitting of the experimental data.
isomerism was also investigated for this molecule, by
scanning particular dihedral angles, in view of determining
the corresponding rotational energy barriers—e.g.
(O12C11C10C9) and (C4C3C9H19), which define the internal
rotation of the carboxylic group and the aromatic ring,
respectively (Fig. 1).
2.3. Raman spectroscopy
The Raman spectra were obtained at room temperature,
in a triple monochromator Jobin–Yvon T64000 Raman
system (0.640 m, f/7.5) with holographic gratings of
1800 grooves.mm21. The detection system was a non-
intensified CCD (Charge Coupled Device) and the entrance
slit was set to 200 mm. The 514.5 nm line of an Arþ laser
(Coherent, model Innova 300) was used as excitation
radiation, providing ca. 70 mW at the sample position.
Samples were sealed in Kimax glass capillary tubes of
0.8 mm inner diameter. Under the above mentioned
conditions, the error in wavenumbers was estimated to be
within 1 cm21. Spectra were collected both for solid
samples and for DMSO-d6 solutions (40 mM).
Twenty one different conformers were found (Fig. 1), the
most stable ones displaying an S-cis orientation of the
terminal carboxylic group. Except for THPPE
5
(DE ¼ 10:9 kJ mol21, out-of-plane H18–(H18O7C6C5) ¼
91.78), THPPE 17 (DE ¼ 31:1 kJ mol21, out-of-plane
H
18 –(H18O7C6C5) ¼ 89.78), THPPE 21 (DE ¼ 37:2 kJ
mol21, out-of-plane H18 –(H18O7C6C5) ¼ 91.58) and
THPPE 18 (DE ¼ 33:2 kJ mol21, out-of-plane H21
–
(H21O13C11C10) ¼ 12.78), all other energy minima have a
planar geometry (Cs symmetry), probably due to the
stabilising effect of p-electron delocalisation between the
ring and the C9yC10 and C11yO12 double bonds, which is
favoured when they are coplanar. In fact, the four most stable
conformers found-THPPE 1 to THPPE 4—are completely
planar ones (Fig. 2). Actually, THPPE 5, which displays a
quasi-perpendicular orientation of the -O7H18 group
((H18O7C6C5) ¼ 91.78), has a much higher relative energy
(DE ¼ 10:9 kJ mol21, Fig. 1).
Fourier transform Raman spectra were recorded on a
RFS 100/S Bruker spectrometer, with a 1808 geometry,
equipped with an InGaAs detector. Near-infrared excitation
was provided by the 1064 nm line of a Nd:YAG laser
(Coherent, model Compass-1064/500 N). A laser power of
200 mW at the sample position was used in all cases, and
resolution was set to 2 cm21
.
The conformational characteristics of THPPE were also
determined to be highly dependent on the geometry of the
carboxylic group 2 either S-cis or S-trans—the former
being strongly favoured: THPPE 1 ðDE ¼ 0Þ vs THPPE 12
(DE ¼ 25:5 kJ mol21), THPPE 2 (DE ¼ 0:5 kJ mol21) vs
THPPE 13 (DE ¼ 28:0 kJ mol21), THPPE 5 (DE ¼
10:9 kJ mol21) vs THPPE 21 (DE ¼ 37:2 kJ mol21), or
THPPE 6 (DE ¼ 11:5 kJ mol21) vs THPPE 19 (DE ¼ 36.3
kJ mol21) (Figs. 1 and 2). Moreover, it was verified that the
orientation of the ring hydroxyl groups has an influence on
the conformation of the carbon pendant chain. In fact, the
geometry corresponding to an identical orientation of these
groups is energetically favoured, once it minimises steric
repulsions between adjacent OH’s, and allows the formation
of medium strength intramolecular H17· · ·O7 and H18· · ·O14,
or H22· · ·O7 and H18· · ·O8 bonds (O· · ·H distances between
217 and 219 pm, for the most stable conformers, Fig. 1). As
2.4. Reagents
3,4,5-trihydroxyaldehyde, trans-caffeic acid, mono-
ethylmalonate and malonic acid were purchased from
´
Sigma-Aldrich Quımica S. A. (Sintra, Portugal). Dimethyl-
sulfoxide-d6 (99.8%) was obtained from E. Merck, Darm-
stadt, Germany. All other reagents and solvents were pro
analysis grade, purchased from Merck (Lisbon, Portugal).
3. Results and discussion
3.1. Ab initio MO calculations
The geometries and relative energies of the distinct
conformers of 3-(3,4,5-trihydro-xyphenyl)-2-propenoic
acid (THPPE) were obtained, through Ab initio MO
calculations (Fig. 1). The effect of several structural
parameters on the overall stability of this molecule was
investigated, namely: (i) S-cis or S-trans conformation of
the carboxylic group-(H21O13C11O12) equal to 08 or 1808,
respectively (atoms numbered according to Fig. 1);
(ii) orientation of the three phenolic hydroxyls (relative to
the ring) 2 dihedrals (H22O14C1C2), (H18O7C6C1) and
(H17O8C5C6) equal to 08 or 1808; (iii) relative orientation
of the aromatic ring and the carboxylic moiety—
(O12C11C10C9) equal to 08 or 1808, and (C11C10C9C3)
equal to 1808 or 08, the former defining a trans or cis
orientation, respectively, of the ring relative to the
carboxylate (around the linear chain CyC bond). Rotational
expected, only high energy geometries (DE $ 16 kJ mol21
)
were obtained when any two ring hydroxyls were directed
towards each other: THPPE 7, THPPE 8, THPPE 9, THPPE
10, THPPE 14 and THPPE 15—displaying an O14· · ·O7
distance of ca. 269 pm-and THPPE 20-with a H18· · ·H17
distance of 196 pm (Fig. 2). As to the relative orientation of
the terminal carboxylic group relative to the phenolic
hydroxyls, THPPE behaves differently from its analogue
caffeic acid (CA) [25], due to the presence of a third OH ring
substituent, which is responsible for a higher symmetry
of the molecule. Thus, as opposed to CA, a syn
conformation (i.e. with CyO and ring OH’s pointing to
the same side relative to the carbon chain) is preferred
over an anti one, either for a (C10C9C3C2) dihedral equal
to 1808 2 THPPE 1 (DE ¼ 0 kJ mol21) vs THPPE 2

106
S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
Fig. 1. Schematic representation of the calculated (B3LYP/6-31G**) conformational energies (and populations, at 25 8C) for 3-(3,4,5-trihydroxyphenyl)-2-
propenoic acid (THPPE). (The atom numbering is included).
(DE ¼ 0:5 kJ mol21)/THPPE 4 (DE ¼ 4:3 kJ mol21), and
THPPE 12 (DE ¼ 25:5 kJ mol21) vs THPPE 13 (DE ¼
vs THPPE 4 (DE ¼ 4:3 kJ mol21), and THPPE 6
(DE ¼ 11:5 kJ mol21) vs THPPE 11 (DE ¼ 20:8kJ mol21
)
(Fig. 2).
Consideration of a cis orientation of the ring and the
carboxylate group relative to the C9yC10 bond 2 (C11C10C9C3)
36:3 kJ mol21)—or 08—THPPE 3 (DE ¼ 3:5 kJ mol21
)

S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
107
and (C10C9C3C2) equal to 08 2 led to rather high energy
conformers: THPPE 6 (DE ¼ 11:5 kJ mol21), THPPE 11
(DE ¼ 20:8 kJ mol21), THPPE 14 (DE ¼ 28:2 kJ mol21),
interactions, yielding a seven-membered intramolecular ring
(Fig. 2), these geometries are not stabilised over the ones
displaying a linear carbon chain, for which p-delocalisation is
surely more effective. THPPE 6 is significantly stabilised as
compared to the other similar conformations, through the
occurrence of a rather strong O1· · ·2H15 interaction
((O12···H15) ¼ 201 pm) coupled to a planar geometry displaying
THPPE
15
(DE ¼ 29:9 kJ mol21),
THPPE
16
(DE ¼ 29:9 kJ mol21), THPPE 17 (DE ¼ 31:1 kJ mol21
)
and THPPE 19 (DE ¼ 36:3 kJ mol21). Actually,
despite the possibility of occurrence of O· · ·H intramolecular
Fig. 2. Representation of the several conformers calculated for 3-(3,4,5-trihydroxy phenyl)-2-propenoic acid (THPPE) 2 displaying (C)H· · ·O and (O)H· · ·O
intramolecular interactions. (B3LYP/6-31G** level of calculation. Distances are represented in pm).

108
S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
Fig. 2 (continued)
an S-cis carboxylate, and a single-directioned orientation of
the ring hydroxyls. The higher energy of THPPE 19, in turn,
is due to the S-trans orientation of its carboxylic group,
which causes a serious steric hindrance between hydrogens H20
and H21 (H20···H21 equal to 203 pm, Fig. 2). Moreover, H-type
bonds between H15 and the carbonyl oxygen (O12) were found
to be preferred over those formed with the hydroxylate O21
atom-THPPE 6 (DE ¼ 11:5 kJ mol21, (O12···H15) ¼ 201 pm)
vs THPPE 16 (DE ¼ 29:9 kJ mol21, (O21···H15) ¼ 212 pm)
(Fig. 2).

S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
Table 1
109
Also, the conformers with a (C10C9C3C2) dihedral
Relative energies and optimised geometries for the most stable conformers
of 3-(3,4,5-trihydroxyphenyl)-2-propenoic acid (THPPE). (B3LYP/6-
31G** level of calculation)
equal to 1808 were found to be favoured relative to the
ones with (C10C9C3C2) ¼ 08, possibly due to both p-
delocalisation and steric repulsion effects-e.g. THPPE 1
(H16· · ·H20 ¼ 223 pm) vs THPPE 3 (H15· · ·H20 ¼ 219 pm)
(Figs. 1 and 2). These conformational preferences are in
accordance with previous results obtained for caffeic
acid [25].
THPPE 1
b0.00; 3.5
THPPE 2
0.45; 4.7
THPPE 3
3.47; 3.3
DE (kJ mol21)/am (D)
Bond lengths (pm)
cC1–C6
C2–C1
139.6
139.0
140.9
145.9
140.7
139.0
140.2
134.7
147.2
137.5
136.1
136.8
121.8
136.1
108.7
108.3
108.9
108.4
96.9
139.8
139.4
140.5
145.9
141.0
138.7
140.1
134.7
147.1
136.1
137.5
136.8
121.7
136.2
108.5
108.6
108.9
108.5
96.9
140.1
138.7
141.0
146.1
140.5
139.4
139.8
134.8
146.9
137.5
136.1
136.8
121.8
136.2
108.6
108.5
108.7
108.5
96.9
Table 1 comprises the calculated optimised geometries
for the lowest energy conformers of THPPE (values for the
other conformers are available from the authors upon
request). These structural parameters do not deviate much
from the X-ray values found in the literature for the
analogous dihydroxylated caffeic acid [43]. Also, they are in
very good agreement with the results obtained from a
complete conformational analysis reported for caffeic acid
[25], as well as with those calculated by Bakalbassis et al.
(at the B3LYP/6-31 þ G* level) for one of caffeic acid’s
ground-state geometries [24].
C3–C2
C3–C9
C4–C3
C5–C4
C6–C5
C9–C10
C
10–C11
C1–O14
C5–O8
C6–O7
C
C
11–O12
11–O13
Potential-energy profiles and energy barriers for internal
rotation around several bonds within the THPPE molecule 2
O7–C6, C3–C9, C10–C11, and C11–O13—were obtained, by
scanning the corresponding torsional angles (Fig. 3).
The (H18O7C6C5) dihedral defines the orientation of the
central ring hydroxyl group relative to the aromatic plane,
i.e. the internal rotation around the O7–C6 bond. The
corresponding potential-energy profile (Fig. 3A), clearly
evidences the preference for a planar geometry, once the
term in 1808 was found to be strongly stabilising—
V11808 ¼ 2 45.9 kJ mol21, as compared to the cosine terms
C2–H15
C4–H16
C9–H19
C
10–H20
O7–H18
O8–H17
96.9
96.5
96.9
O
13–H21
97.2
97.2
97.1
O14–H22
96.5
96.9
96.5
Bond angles (8)
C2–C3–C4
119.2
118.0
128.0
119.8
119.9
111.5
122.0
107.9
115.9
123.3
105.6
109.9
119.2
118.2
128.0
120.0
114.3
111.5
122.0
109.9
116.1
123.3
105.6
107.9
119.2
122.7
127.5
124.0
120.3
114.0
121.8
107.8
117.3
122.8
105.3
109.9
C9–C3–C2
in 908 and 608—V2908 ¼ 2 0.7 kJ mol21 and V3608
¼
C10–C9–C3
C11–C10–C9
O8–C5–C6
2 3.3 kJ mol21, respectively. Also, it was verified that the
most unfavoured geometries correspond to (H18O7C6C5)
equal to either 08 or 3608, which is easily understandable in
view of the significant steric hindrance between H18 and H17
(H18· · ·H17 equal to 149 pm, O7· · ·O14 being 272 pm). In
fact, due to the planarity of the OH groups relative to the
ring (imposed during the scanning process), the H18· · ·H17
repulsion is much more pronounced than, for instance, in
conformer THPPE 20, where the H18· · ·H17 distance is equal
to 196 pm and a O7· · ·H22 intramolecular bond can be
formed ((O7· · ·H22) ¼ 209 pm, Fig. 2). Furthermore, values
of either 1808 or 08 for the three (H18O7C6C1), (H17O8C5C4)
or (H22O14C1C6) dihedral angles, simultaneously, yield very
stable conformations (e.g. THPPE 1 and THPPE 2, Fig. 2),
where H18· · ·O14/H17· · ·O7, and H22· · ·O7/H18· · ·O8 intra-
molecular interactions may occur (H· · ·O varying from 217
to 219 pm). For different values of those dihedrals, however,
the strong steric hindrance between atoms H17 and H18, and/
or H18 and H22, leads to a pronounced deviation from
planarity of some of the ring hydroxyl groups (e.g. THPPE
17, (H18O7C6C5) ¼ 293.98, Fig. 2). The internal rotational
energy barrier from (H18O7C6C5) ¼ 180–08 (or 3608) was
thus calculated to be rather high 249.1 kJ mol21. The
scanning of dihedral (H18O7C6C5) was carried out for a
particular conformation of the two neighbouring OH
O13–C11–C10
O
13–C11–O12
H17–O8–C5
19–C9–C10
H
H20–C10–C9
H21–O13–C11
H
22–O14–C1
Dihedral angles (8)
C3–C4–C5–C6
C9–C3–C4–C5
0.0
2180.0
2180.0
180.0
0.0
0.0
180.0
2180.0
180.0
0.0
0.0
180.0
0.0
C
C
10–C9–C3–C2
11–C10–C9–C3
180.0
180.0
0.0
O12–C11–C10–C9
O
H
13–C11–C10–C9
15–C2–C3–C9
2180.0
0.0
180.0
0.0
0.0
H16–C4–C3–C2
180.0
0.0
180.0
2180.0
2180.0
0.0
180.0
0.0
H
H
H
17–O8–C5–C6
18–O7–C6–C1
20–C10–C9–C3
0.0
0.0
0.0
0.0
H21–O13–C11–C10
H22–O14–C1–C2
180.0
0.0
2180.0
180.0
180.0
0.0
a
Total dipolar moment 1D ¼ 1/3 £ 1022 Cm.
b
Total value of energy for the most stable conformer of THPPE is
2 723.904147480 (in Hartree, 1 Hartree ¼ 2625.5001 kJ mol21).
c
Atoms are numbered according to Fig. 1.

110
S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
groups—(H17O8C5C4) and (H22O14C1C6) equal to 1808—
once, as expected, variation of the orientation of one of them
affects the position of the others. This conformational
behaviour of the ring hydroxyls in THPPE differs from the
one encountered for caffeic acid [25], on account of the
presence of three OH ring substituents in the latter, in a
symmetric position relative to the pendant carbon chain.
This explains the ocurrence of a maximum energy geometry
for (H18O7C6C5) ¼ 0/3608, as opposed to caffeic acid, for
which the absence of the third hydroxyl allows a certain
stabilisation in these conditions, the most unfavourable
conformation corresponding to (H18O7C6C5) ¼ 908
(V11808 ¼ 245.9 vs 215.1 kJ mol21, and V9208 ¼ 20.7 vs
22.3 kJ mol21, respectively, for THPPE and CA). Actually,
the presence of three OH ring substituents is prone to affect
the electronic delocalisation process within the THPPE
molecule, as compared to the dihydroxylated analogue.
The orientation of the pendant carbon chain relative
to the aromatic ring is defined by dihedral (C10C9C3C2)—
e.g. THPPE 1 ((C10C9C3C2) ¼ 08) and THPPE 7
Fig. 3. Optimised conformational energy profiles, and their Fourier deconvolutions, for several internal rotations within the 3-(3,4,5-trihydroxyphenyl)-2-
propenoic acid (THPPE) molecule. A 2 around the O7–C6 bond: V1 ¼ 245.9 kJ mol21, V2 ¼ 20.7 kJ mol21, V3 ¼ 23.3 kJ mol21. B—around the C9–C3
bond: V1 ¼ 21.2 kJ mol21, V2 ¼26.0 kJ mol21 and V3 ¼ 20.7 kJ mol21. C—around the C11–C10 bond: V1 ¼1.9 kJ mol21, V2 ¼ 35.6 kJ mol21. D—around
the O13–C11 bond: V1 ¼ 22.8 kJ mol21, V2 ¼ 43.2 kJ mol21. (When varying (H18O7C6C5), both (H22O14C1C6) and (H17O8C5C4) were frozen at 1808.
B3LYP/6-31G* level of calculation).

S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
111
Fig. 3 (continued)
((C10C9C3C2) ¼ 1808). The analysis of the Fourier com-
ponents of the corresponding potential-energy variation
(Fig. 3B) evidences the preference for a planar geometry,
i.e. for a completely conjugated system—in fact, the cosine
term in 908 ðV2Þ is by far the primary positive contribution
21.2 kJ mol21 and V6308 ¼ 20.7 kJ mol21). This may be
explained by the strong stabilisation due to p-electron
delocalisation, which is considerably more effective for a
linear zig–zag like unsaturated pendant chain, coplanar
with the aromatic ring. A value of 27.2 kJ mol21 for the
internal rotational energy barrier—from (C10C9C3C2) equal
to 180–908 was obtained.
around the C10–C11 bond—e.g. geometries THPPE 1
((O12C11C10C9) ¼ 08) and THPPE 4 ((O12C11C10C9) ¼
1808). The term in 908 was found to be the ruling
unstabilising contribution (V9208 ¼ 35.6 kJ mol21 vs
V11808 ¼ 1.9 kJ mol21), the planar arrangement of the
carboxylate group relative to the aromatic ring and the
carbon chain double bond being once more highly favoured
(DE ¼ 39:9 kJ mol21).
to that profile (V9208 ¼ 26.0 kJ mol21 vs V11808
¼
Rotation around the C11–O13 bond, defining either an
S-cis or S-trans conformation within the carboxylic group,
is represented in Fig. 3D, and corresponds to the highest
energy barrier calculated for this molecule (DE ¼
56:8 kJ mol21). A clear preference for an S-cis orientation
is obtained, the less favoured conformation being the
one with (H21O13C11C10) ¼ 908, due to the loss of
The variation of the (O12C11C10C9) dihedral (Fig. 3C)
represents the internal rotation of the carboxylate group,

112
S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
planarity of the system—V9208 ¼ 43.2 kJ mol21 vs
V11808 ¼ 222.8 kJ mol21. In fact, the geometry displaying
a 908 orientation of the carboxylic OH goup relative to the
plane containing both the carbon chain and the ring
corresponds to a maximum in the potential-energy profile,
and not the S-trans geometry. Actually, the latter
corresponds to a planar arrangement of the molecule,
despite some steric hindrance which may occur between
atoms H20 and H21 (e.g. H20· · ·H21 ca. 215 pm in THPPE
12 and 13, Fig. 2). It is thus concluded that the most
stable conformations of the carboxylate moiety are
the ones corresponding to (O12C11C10C9) ¼ 08 and
from ca. 800 and 1000 cm21, typical of tri-substituted
benzenes; (iv) CH deformations—both wagging, at ca.
1370 cm21, and rocking modes, at 700–870 cm21; (v) CH
characteristic stretching vibrations, both symmetric and
assymmetric, respectively, around 3050–3125 cm21
;
(vi) bands around 1290 and 1150–60 cm21, which are
assigned, in THPPE, to both nCO and dOH vibrations: while in
the acid species these are broad features, clearly comprising
more than one mode, in the ester they are narrower, probably
corresponding solely to the OH deformations.
The nOH modes—either from the carboxylic group or the
hydroxyl ring substituents—were not detected experimen-
tally for THPPE, while the ones observed for both CA and
ETHPPE displayed a significant shift relative to the
predicted values (Table 2). In fact, formation of stabilising
(O)H· · ·O(yC) intermolecular close contacts in this kind of
phenolic compounds is known to occur, yielding rather
stable dimeric species, and giving rise to downward
frequency shifts of both the nCyO and nOH modes (as
previously observed for substituted benzaldehydes [45,46]).
Raman spectra of THPPE in solution was thus obtained, in
view of better understanding the behaviour of the carboxylic
nCyO oscillator upon dilution (e.g. decrease of the
dimer:monomer ratio). DMSO-d6 was chosen as a suitable
solvent for this purpose, as it does not display any bands in
the spectral region of interest. Moreover, any interaction
that might occur between the THPPE molecule and
DMSO—via (O)H· · ·O(yS) close contacts—was detected
only for very high dilutions, through a band appearing at ca.
1750 cm21 [47], which did not interfere with the features
under study. Two bands ascribed to the CyO stretching
mode—at both 1692 and 1636 cm21—were detected for the
THPPE-DMSO-d6 solution (40 mM), as opposed to the
solid, for which only the feature at lower wavenumber
(1640 cm21) was to be seen with significant intensity
(Table 2, Fig. 6). These two bands are suggested to be due to
the free and hydrogen-bonded forms of the CyO group,
respectively. Actually, while the dimeric species is expected
to be predominant in the solid (nCyO ¼ 1640 cm21),
dilution may be responsible for the presence of the
monomeric form, which is reflected by the observation of
free carboxylate groups (not involved in intermolecular H-
bonds)—nCyO vibration (1692 cm21). The higher electronic
delocalisation occuring in THPPE (due to the presence of
the linear chain double bond) relative to the aromatic
aldehydes previously reported [45,46] leads to a larger shift
(ca. 50 cm21) then the one measured for benzaldehyde
dimers (ca. 10 cm21). Also, as expected for this kind of
dimeric–monomeric carboxyl modes, the C–O stretching
(H21O13C11C10) ¼ 1808—THPPE 1, THPPE
2 and
THPPE 3.
The harmonic vibrational frequencies were calculated for
all conformers of THPPE (data available from the authors
upon request). The ones for the most stable geometries are
comprised in Table 2 and show a good overall agreement
with both the experimental data (after scaling according to
Scott and Radom for the particular basis set used [40]), and
the theoretical values previously obtained by the authors for
caffeic acid [25].
In order to better mimetise THPPE in the solid state, and
once this kind of phenolic carboxylic acids predominantly
occur (in the condensed phase) as hydrogen-bonded dimers,
calculations were also carried out for a dimeric species,
formed through intermolecular top-to-top (O)H· · ·O(yC)
interactions between the carboxylate groups of each
molecule (Fig. 4, Table 2). This dimeric structure was
found to be energetically favoured, displaying quite strong
hydrogen close contacts-(O)H· · ·O(yC) distances of 162
and 163 pm.
3.2. Raman spectroscopy
Fig. 5 represents the solid state Raman spectra (at room
temperature) of 3-(3,4,5-trihydroxyphenyl)-2-propenoic
acid and its ethyl ester (ETHPPE), in the 150 to
1750 cm21 region. The corresponding experimental
Raman wavenumbers are gathered in Table 2, along with
the values calculated (at the B3LYP/6-31G** level) for the
two most stable conformers of THPPE and for a dimeric
form of this molecule (represented in Fig. 4). Although the
calculated spectrum of the dimer contains twice the number
of frequencies compared to the monomer, only those
presenting a better agreement with the experimental values
are comprised in Table 2.
A complete assignment of the Raman bands of THPPE
was performed, based on the theoretical results presently
obtained for this molecule, as well as on previous studies on
similar phenolic compounds [25,44] (Table 2). The main
observed features were (Fig. 5): (i) CyC double bond
stretching vibrations—at ca. 1640 and 1612 cm21 (1661 and
1601 cm21, respectively, in the ester); (ii) aromatic ring nCC
mode-at 1583 and 1612 cm21 (1601 cm21 in the ester);
(iii) out-of-plane ring wagging vibrations—at wavenumbers
vibration was found to be shifted from 1289 to 1267 cm21
,
when going from the solid state to the solution (Fig. 6). The
OH bending vibration, detected at about 1355 cm21 in the
solution, is also typical of monomeric forms, while the
corresponding mode for dimers appears at slightly higher
values (at 1374 cm21). The band appearing at 1462 cm21 in
the spectrum of the solution, which is assigned to both C–O

Table 2
Experimental (solid state) and calculated (B3LYP/6-31G**) Raman wavenumbers (cm21) for the most stable conformers of 3-(3,4,5-trihydroxyphenyl)-2-propenoic acid (THPPE), as well as for a THPPE 1
dimeric structure. (Values for CA and ETHPPE are included for comparison)
Experimental
THPPE
Calculateda
Approximate descriptionb
CA
ETHPPE
THPPE 1c(44%)
THPPE 2 (37%)
THPPE 1-Dimer
CAd
3696 (72;117)
3648 (140;180)
3637 (99;100)
3619 (87;198)
3101 (3;55)
3695 (64;98)
3646 (126;159)
3640 (101;108)
3619 (88;196)
3091 (2;65)
3696 (71;117)
3648 (110;280)
3637 (76;146)
3693 (61; 98)
n (O14H22
)
)
3368
3341
3321
3370
3341
3321
3627 (123;177)
n (O7H18
n (O8H17
)
)
3619 (89;193)
3095 (6;169)
3084 (11;36)
3073 (5;62)
3058 (8;25)
3049 (1;31)
n (O13H21
3124
3080
3067
3053
3102 (2;52)
3091 (0;24)
3058 (6;88)
3048 (2;22)
n (CH)ring þ þ n (CH)chain
3084
3079
3040
3033
3089 (4;14)
3084 (10;48)
3068 (4;22)
n (CH)chain
n (CH)ring
3064
3035
3059 (8;71)
3045 (1;27)
3051 (1;34)
n (CH)chain
n (CH)
2879 (2;2879)
n (O13H21
n (CH)
)
3026
2992
3020
(1289 þ 1612/1640) cm21
3009
2988
2966
2942
2899
n (CH)
1697
1640
1612
1739 (258;85)
1634 (218;659)
1742 (266;92)
1631 (239;704)
1690 (481;0)
1646 (0;664)
1629 (1;1482)
1601 (0;4433)
1599 (1;1121)
1741 (270;87)
1630 (208;699)
n (CyO)
n (CyC)
1640
1613
1660
1602
n (CyC) þ n (CyO) þ n (CC)ring
n (CC)ringþn (CyC)chain
n (CC)ring
1601 (157;1606)
1599 (157; 1606)
1605 (75;509)
1599 (271;1481)
1584 (29;22)
1583
1595
1532
1525
1597 (194;1052)
1512
1476
1517 (133;18)
1459 (232;3)
1518 (209;1)
1455 (81;14)
1517 (0;1)
1512 (178;2)
n (CC)ring
1436 (177;144)
n (CC)ring þd (O8H17
)
1464 (0;164)
1457 (0;710)
d (O8H17) þ d (O13H21) þ d (CH)chain
d (OH)
1374
1355
1386
1375 (22;96)
1378 (50;125)
1379 (128;35)
n (CC)ring þ d (OH)ring þ d (CH)chain
1365
1358
n (CC)ring þ d (O8H17) þ d (O13H21) þ d (CH)chain
n (CC)ring þ d (O8H17) þ d (O13H21) þ d (CH)chain
n (CC)ring þd (O7H18) þ d (CH)
1354
1316
1299
1286
1362 (1;99)
1354 (174;146)
1339 (24;2)
1344 (31;2)
1314 (0;43)
1295 (18;6)
1277 (379;32)
1250 (10;94)
1223 (4;7)
1316
1282
1329 (33;39)
1306 (157;2)
1292 (428;140)
1261 (178;46)
1221 (118;18)
1212 (17;20)
1177 (84;2)
d (O8H17) þ d (O13H21) þ d (CH)chain
1307
1289
1245
1236
1204
1151
1141
1308 (167;11)
1275 (280;80)
1270 (128;28)
1221 (74;11)
1214 (85;11)
1170 (95;4)
n (CC)ring þn (CO) þ d (CH) þ d (OH)ring þ d (OH)carbox
d (OH)ring þ d (OH)carbox þ d (CH)
1311 (0,82)
1296 (961;9)
n (CC)ring þn (CO) þ d (CH) þ d (OH)ring þ d (OH)carbox
n (CC)ring þn (CO) þ d (CH) þ d (CH) þ d (OH)ring þ d (OH)carbox
1186
1177
1204
1151
1173 (145;118)
1147 (5;47)
1129 (129;7)
d (CH) þ n (C5O8) þ d (O7H18) þ d (O8H17) þ d (O13H21
d (O7H18) þ n (CH) þ n (CO)
d (O13H21) þ d (O14H22) þ d (CH)
d (OH)ring þ d (CH)
)
1138 (120;5)
1135 (36;51)
1136 (0;541)
1131 (24;97)
d (OH)ring þ d (CH)
(continued on next page)

Table 2 (continued)
Experimental
Calculateda
Approximate descriptionb
THPPE
CA
ETHPPE
THPPE 1c(44%)
THPPE 2 (37%)
THPPE 1-Dimer
CAd
1122
1006
1122
1006
1132 (3;60)
1105 (103;3)
1015 (573;284)
991 (198;9)
974 (6;6)
1130 (39;1)
1103 (665; 280)
1014 (149;3)
991 (22;2)
1108 (625;201)
1091 (128;30)
992 (21;3)
d (O7H18) þ d (O14H22) þ d (CH)ring
d (O13H21) þ d (O14H22) þ d (CH)
n (CC)ring þ n (CO)ring d (CH) þ d (OH)ring
g (CH)chain
1005
992
1015 (46;18)
990 (14;3)
976 (0;10)
957 (15;8)
972
976 (5;8)
930 (15;6)
n (CC)ring þ d (CCC)chain
935 (16;9)
935 (32;11)
907 (1;3)
n (CC)ring þ n (CC)ring þ d (OH)
925 (0;2)
g (O13H21)
d (CCC)chain
g (C10H20
850 (0;36)
868
810
874
854
812
872
854
811
847 (12;13)
819 (12;2)
847 (4;12)
822 (30;2)
843 (5;10)
811 (48;1)
)
g (C4H16
g (CH)ring
d (CCC)
)
820 (5;3)
784 (0;52)
782 (20;19)
775 (41;1)
721 (24;2)
678 (2;3)
782 (11;21)
765 (29;2)
721 (27;2)
679 (4;3)
796 (14;6)
785 (5;29)
747 (9;8)
n (CC)ring þ n (CO)ring
761
741
g (C2H15)
g (C2H16) þ g (CCC)chain
n (CC)ring þ d (CCC)chain
739
671
665
624
605
589
740
696 (0;4)
721 (24;3)
647
622
647
623
616 (11;3)
608 (61;6)
581 (82;12)
558 (26;2)
614 (39;2)
608 (65;6)
581 (56;14)
559 (26;2)
619 (0;17)
625 (22;13)
603 (69;4)
579 (28;5)
560 (26;4)
528 (46;1)
498 (12;1)
454 (76;3)
439 (13;3)
d (CCC)chain
g (O13H21) þ G (CCC)
D (CCC)ring þ D (CCC)chain
D (CCC)ring þ D (CCC)chain þ g (O13H21
)
541
521
484
452
434
380
352
323
290
541
541
502
481
D (CCC)ring þ g (O13H21
)
522 (0;4)
521 (1;4)
489 (32;6)
454 (4;2)
415 (39;2)
386 (16;4)
360 (123;3)
316 (12;1)
296 (1;1)
277 (1;2)
250 (5;4)
214 (174;2)
213 (2;4)
161 (0;1)
524 (0;9)
499 (0;26)
462 (0;6)
423 (42;2)
388 (0;8)
357 (117;2)
301 (0;5)
292 (0;17)
274 (0;3)
252 (0;2)
D (CCC)ring þ d (CCC)chain
D (CCC)
D (CCC)
489 (25;6)
454 (11;3)
426 (51;1)
387 (20;5)
357 (104;2)
316 (10;1)
g (O8H17) þ g (O7H18
)
395
352
381 (1;2)
G (CCC)
g (O14H22) þ g (O8H17) þ g (O7H18
D (COH)ring
D (COH)ring
)
290 (4;2)
247 (10;4)
205 (2;5)
–
274
262
229
276 (2;4)
251 (7;1)
213 (0;3)
209 (179;5)
162 (0;1)
D (CCC)
G (CCC)
D (CCC)chain
264
241
207 (49;8)
176 (0;7)
163 (0;8)
216 (139;1)
187 (15;0)
g (O7H18) þ g (O14H22
)
179
177
141
G (CCC)
D (CCC)
137 (1;1)
135 (1;1)
134 (0;2)
37 (0;10)
33 (0;1)
skeletal modes
skeletal modes
skeletal modes
a
Wavenumbers above 400 cm21 are scaled by a factor of 0.9614 [40]. (IR intensities in km mol21; Raman scattering activities in A.amu).
˚
b
c
d
Atoms are numbered according to Fig. 1.
Relative population (at 25 8C).
[25].

S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
115
Fig. 4. Representation of a calculated dimeric structure for 3-(3,4,5-trihydroxy-phenyl)-2-propenoic acid (THPPE). (B3LYP/6-31G** level of calculation.
Distances are represented in pm).
stretching and in-plane C–O–H deformation 2 thus being
directly affected by the occurrence of dimeric structures
(Fig. 4)-is not observed in the condensed phase. As to the
aromatic ring nCyC mode, it suffers a slight downward shift
when going from the dimer (1612 cm21) to the monomer
(1605 cm21). An identical behaviour was observed for
caffeic acid (CA), the dihydroxylated analogue of THPPE,
which is well known for forming dimeric structures in the
solid phase [48,49].
the formation of intermolecular H-bonds involving those
hydroxyl groups. These findings are corroborated by the Ab
initio results obtained for a THPPE dimeric structure. As
expected, the calculated wavenumbers associated to the
groups involved in intermolecular hydrogen close-contacts
within the dimer (e.g. carboxylic CyO) presented a better
accordance with the experimental values obtained for the
solid than the ones calculated for the isolated molecule. In
fact, both nCyO experimental bands—at 1697 and
1640 cm21-were predicted by the calculations when
Furthermore, the fact that no experimental nOH Raman
bands were seen in the spectrum of solid THPPE is
characteristic of the presence of dimeric species, for which
these vibrations yield extremely broad features at around
3300 cm21, often not detected, as a consequence of
considering the dimer—at 1690 cm21 and 1646 cm21
,
respectively—but not when considering the monomer
(Table 2). Moreover, characteristic bands of the dimeric
species were obtained by the calculations, namely at
Fig. 5. Experimental Raman spectra (150-1750 cm21) of 3-(3,4,5-trihydroxyphenyl)-2-propenoic acid (THPPE) and the corresponding ethyl ester (ETHPPE).
(Solid state, at 25 8C).

116
S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
Fig. 6. Experimental Raman spectra (1100–1750 cm21, at 25 8C) of 3-(3,4,5-trihydroxyphenyl)-2-propenoic acid (THPPE) and the analogous dihydroxylated
caffeic acid (CA), both in the solid state and in DMSO-d6 solution (40 mM). (Solid line-solid state; dotted line-solution).
2879 cm21—assigned to nOH involved in (O)H· · ·O(yC)
and/or (O)H· · ·O(H) intermolecular bonds—and at 1464,
1457 and 925 cm21—ascribed to in-plane and out-of-plane
OH bending modes (Table 2). The similarity between the
Raman pattern obtained for the ETHPPE ester and the
THPPE solution (except for the CyO stretching mode,
Fig. 5), is good evidence of the validity of the previously
discussed assignment for the phenolic acid—two distinct
signals for nCyO in the solid, due to the monomer and dimer
species.
4. Conclusions
The conformational analysis performed for THPPE
rendered twenty one distinct conformers, with structural
differences concerning the conformation of the terminal
carboxylate group and the pendant carbon chain, as well as
the orientation of the three ring hydroxyl substituents. Two
of these conformers—THPPE1 (DE ¼ 0.0 kJ mol21) and
THPPE2 (DE ¼ 0.5 kJ mol21)—were found to be signifi-
cantly more stable, with populations at room temperature of
44 and 37%, respectively. It was verified that the lowest
energy species displayed a planar geometry (CS symmetry),
as well as a carboxylate S-cis conformation. In fact, these
two factors—carboxylate terminal group orientation and
coplanarity of the molecule—appear to be the most
The agreement between the results now obtained for
THPPE—both calculated and experimental vibrational
wavenumbers—and the data found in the literature for
similar molecules, both from theoretical and spectroscopic
studies [25,44,48–50], was found to be quite good.

S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
117
[9] H. Esterbauer, J. Gebicki, H. Puhl, G. Jurgens, Free Rad. Biol. Med.
13 (1992) 341.
significant stabilising factors in this kind of systems: The
preference for planarity is predictable, once it favours
electron delocalisation through the expanded p system of
the hyperconjugated molecule of THPPE. Thus, non-planar
geometries arose only in order to overcome steric hindrance
destabilising factors, whose minimisation resulted to be
more favorable then the maintenance of planarity. Potential-
energy profiles for internal rotation around different bonds
within the molecule also supported the energetically
favoured planar arrangement.
[10] B. Halliwell, J.C. Gutteridge, Free Radicals in Biology and Medicine,
Oxford Science Publications, 1999.
[11] S. Passi, M. Picardo, M. Nazzaro-Porro, Biochem. J. 245 (1987)
537.
[12] T. Nakayama, Cancer Res. 54 (Suppl) (1994) 1991.
[13] E. Sergediene, K. Jonsson, H. Szymusiak, B. Tyrakowska, I.M.
Rietjens, N. Cenas, FEBS Lett. 462 (1999) 392.
[14] M. Inoue, N. Sakaguchi, K. Isuzugawa, H. Tani, Y. Ogihara, Biol.
Pharm. Bull. 23 (2000) 1153 and references therein.
´ ´
[15] G. Roy, M. Lombardıa, C. Palacios, A. Serrano, C. Cespon, E. Ortega,
´ ´
P. Eiras, S. Lujan, Y. Revilla, P. Gonzalez-Porque, Arch. Biochem.
The Raman data presently obtained for THPPE is in total
agreement with previous studies on similar phenolic
compounds [25,44]. The presence of dimeric species
in the solid state (predicted by the calculations) was clearly
evidenced in the Raman pattern obtained for both
the condensed phase and the solution, mainly through the
characteristic low frequency shifts detected for both the
nCyO and nC–O vibrational modes.
Biophys. 383 (2000) 206 and references therein.
[16] T. Gao, Y. Ci, H. Jian, C. An, Vib. Spec. 24 (2000) 225.
[17] C.A. Gomes, T. Gira˜o da Cruz, J.L. Andrade, N. Milhazes, F. Borges,
M.P.M. Marques, J. Med. Chem. 46 (2003) 5395 (DOI: http://dx.doi.
org/10.1021/jm030956v).
[18] F.A.M. Silva, F. Borges, C. Guimara˜es, J.L.F.C. Lima, C. Matos, S.
Reis, J. Agric. Food Chem. 48 (2000) 2122.
[19] F. Natella, M. Nardini, M. Di Felice, C. Scaccini, J. Agric. Food
Chem. 47 (1999) 1453.
In conclusion, this kind of vibrational spectroscopic
studies carried out for phenolic acids, coupled to Ab initio
theoretical calculations, are of the utmost importance for
understanding the structure-activity relationships ruling the
biological activity of such compounds, which, apart from
being recognised antioxidants, often reveal cytotoxic
properties against human cancer cell lines. Actually,
cytotoxic and growth-inhibition assays with both THPPE
and analogous compounds displaying potential anti-
cancer activity (e.g. caffeic and gallic acids, as well as
some of their esters) are presently being performed in
our laboratory [17], and can only be accurately interpreted in
the light of the corresponding conformational analysis.
[20] K. Tanaka, S. Sakai, S. Tomiyama, T. Nishiyama, F. Yamada, Bull.
Chem. Soc. Jpn 64 (1991) 2677.
[21] S. Tomiyama, S. Sakai, T. Nishiyama, F. Yamada, Bull. Chem. Soc.
Jpn 66 (1993) 299.
[22] I. Vedernikova, E. Proinov, D.R. Salahub, A. Haemers, Int. J. Quant.
Chem. 77 (1999) 161.
[23] P. Rajan, I. Vedernikova, P. Cos, D. Vanden, K. Berghe, A.
Augustyns, A. Haemers, Bioorg. Med. Chem. Lett. 11 (2001) 215.
[24] E. Bakalbassis, A. Chatzopoulou, V.S. Melissas, M. Tsimidou, M.
Tsolaki, A. Vafiadis, Lipids 36 (2001) 181.
[25] E. van Besien, M.P.M. Marques, J. Mol. Struct. (THEOCHEM) 625
(2003) 265.
[26] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A.
Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery Jr.,
R.E. Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D.
Daniels, K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone,
M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo,
S. Clifford, J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K.
Morokuma, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B.
Foresman, J. Cioslowski, J.V. Ortiz, A.G. Baboul, B.B. Stefanov,
G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L.
Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng,
A. Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W.
Chen, M.W. Wong, J.L. Andres, C. Gonzalez, M. Head-Gordon,
E.S. Replogle, J.A. Pople, GAUSSIAN 98, Revision A.9, Gaussian
Inc., Pittsburgh, PA, USA, 1998.
Acknowledgements
SF and MPM thank the Chemistry Department of the
University of Aveiro (in the person of Dr Helena Nogueira),
where the FT-Raman experiments were carried out. NM
is grateful to FCT for a fellowship (PRAXIS XXI/BD/
18520/98).
[27] T.V. Russo, R.L. Martin, P.J. Hay, J. Phys. Chem. 99 (1995) 17085.
[28] A. Ignaczak, J.A.N.F. Gomes, Chem. Phys. Lett. 257 (1996) 609.
[29] F.A. Cotton, X. Feng, J. Am. Chem. Soc. 119 (1997) 7514.
[30] A. Ignaczak, J.A.N.F. Gomes, J. Electroanal. Chem. 420 (1997)
209.
References
[1] O.I. Aruoma, A. Murcia, J. Butler, B. Halliwell, J. Agric. Food Chem.
41 (1993) 1880.
[31] T. Wagener, G. Frenking, Inorg. Chem. 37 (1998) 1805.
[32] F.A. Cotton, X. Feng, J. Am. Chem. Soc. 120 (1998) 3387.
[33] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B37 (1988) 785.
[34] B. Miehlich, A. Savin, H. Stoll, H. Preuss, Chem. Phys. Lett. 157
(1989) 200.
[2] C.A. Rice-Evans, N.J. Miller, G. Paganga, Free Rad. Biol. Med. 20
(1996) 933.
[3] G. Cao, E. Sofic, R.L. Prior, Free Rad. Biol. Med. 22 (1997) 749.
[4] V.L. Singleton, Am. J. Enol. 38 (1987) 69.
[5] P. Groupy, A. Fleuriet, M.J. Amiot, J.J. Macheix, J. Agric. Food
Chem. 39 (1991) 92.
[35] A. Becke, Phys. Rev. A38 (1988) 3098.
[36] A. Becke, J. Chem. Phys. 98 (1993) 5648.
[37] P.C. Hariharan, J.A. Pople, Theor. Chim. Acta 28 (1973) 213.
[38] M.M. Francl, W.J. Pietro, W.J. Hehre, J.S. Binkley, M.S. Gordon, D.J.
DeFrees, J.A. Pople, J. Chem. Phys. 77 (1982) 3654.
´ ´
[6] M. Brenes-Balbuena, P. Garcıa-Garcıa, A. Garrido-Fernandez,
J. Agric. Food. Chem. 40 (1992) 1192.
´
[7] A. Serrano, C. Palacios, G. Roy, C. Cespon, M.L. Villar, M. Nocito,
[39] C. Peng, P.Y. Ayala, H.B. Schlegel, M.J. Frisch, Comp. Chem. 17
(1996) 49.
´
´
P. Gonzalez-Porque, Arch. Biochem. Biophys. 350 (1) (1998) 49.
[8] F.A.M. Silva, F. Borges, M.A. Ferreira, J. Agric. Food Chem. 49
(2001) 3936.
[40] A.P. Scott, L. Radom, J. Phys. Chem. 100 (1996) 16502.
[41] L. Radom, W.J. Hehre, J.A. Pople, J. Am. Chem. Soc. 94 (1972) 2371.

118
S.M. Fiuza et al. / Journal of Molecular Structure 693 (2004) 103–118
[42] L.A.E. Batista de CarvalhoT, A.M. Amorim da Costa, J.J.C. Teixeira-
Dias, J. Mol. Struct. (Theochem) 205 (1990) 327.
[46] M.P.M. Marques, A.M. Amorim da Costa, P.J.A. Ribeiro-Claro,
J. Phys. Chem. A 105 (2001) 5292.
[47] P. Novak, D. Vikic-Topic, Z. Meic, S. Sekusak, A. Sabljic, J. Mol.
Struct. 356 (1995) 131.
´ ´ ´
[48] S. Sanchez-Cortes, J.V. Garcıa-Ramos, Spectrochim. Acta A55
´
[43] S. Garcıa-Granda, G. Beurskens, P.T. Beurskens, T.S.R. Krishna,
G.R. Desiraju, Acta Cryst. C43 (1987) 683.
[44] R. Calheiros, N. Milhazes, F. Borges, M.P.M. Marques, J. Mol. Struct.
(2004) in press.
(1999) 2935.
´ ´
[49] S. Sanchez-Cortes, J.V. Garcıa-Ramos, Appl. Spec. 54 (2000) 230.
´
[45] N. Karger, A.M. Amorim da Costa, P.J.A. Ribeiro-Claro, J. Phys.
Chem. A 103 (1999) 8672.
´ ´ ´
[50] S. Sanchez-Cortes, J.V. Garcıa-Ramos, J. Colloid, Interface Sci. 231
(2000) 98.