Changes in the aggregation patterns of Z-2,3-diphenylpropenoic acid and its methyl ester on substituting the olefinic hydrogen with CF3 group - An FT-IR study

Article abstract of DOI:10.1016/j.molstruc.2004.11.034

While in the unsubstituted Z-2,3-diphenylpropenoic acid and its methyl ester the olefinic protons rarely were part of any hydrogen bonding interaction, upon substitution by CF₃ group, the possibility of (aromatic)C-H?F intermolecular hydrogen b

Full text of DOI:10.1016/j.molstruc.2004.11.034

Journal of Molecular Structure 744–747 (2005) 207–210
www.elsevier.com/locate/molstruc
Changes in the aggregation patterns of Z-2,3-diphenylpropenoic acid
and its methyl ester on substituting the olefinic hydrogen
with CF3 group—an FT-IR study
*
´
´
¨
J.T. Kiss, K. Felfoldi, I. Palinko
´ ´
Department of Organic Chemistry, University of Szeged, Dom ter 8, Szeged H-6720, Hungary
Received 6 September 2004; revised 9 November 2004; accepted 9 November 2004
Available online 11 January 2005
Abstract
While in the unsubstituted Z-2,3-diphenylpropenoic acid and its methyl ester the olefinic protons rarely were part of any hydrogen bonding
interaction, upon substitution by CF3 group, the possibility of (aromatic)C–H/F intermolecular hydrogen bond appeared and indeed realised
for the molecules (E-2,3-diphenyl-3-CF3-propenoic acid and its methyl ester) in the solid state. This type of close contact was indicated
experimentally by FT-IR spectroscopy.
q 2004 Elsevier B.V. All rights reserved.
Keywords: E-2,3-Diphenyl-3-CF3-propenoic acid and its methyl ester; The effect of CF3 substituent; FT-IR spectroscopy; C–H/F hydrogen bonds
1. Introduction
2. Experimental
Beside dimerisation, typical for carboxylic acids, a-
phenylcinnamic acid (2,3-diphenylpropenoic acid) stereo-
isomers and their derivatives do form extended structures
kept together by weak C–H/O hydrogen bonds in the solid
state [1,2]. These hydrogen bonds are of two types. Either
the carbon, playing the role of hydrogen donor, is part of one
of the aromatic rings [1,2] or it is the appropriate olefinic
pillar atom [2]. The latter type of interaction is rare; it only
appears in the methyl esters. Possible reason of its non-
existence in the acids is the steric hindrance between the
phenyl groups arising during dimerisation. This hindrance
may be effectively decreased by substituting the small
olefinic hydrogen by a large CF3 group. This way extra
hydrogen bond accepting atoms are built in the molecule
and they are farther away from the molecular skeleton
than the hydrogen atom was. During experimental work
the intermolecular hydrogen bonding patterns of these
CF3-substituted molecules were investigated. Results of this
study are communicated in this contribution.
The E-2,3-diphenyl-3-propenoic acid (1) (a derivative of
Z-2,3-diphenyl propenoic acid) and its methyl ester (2) were
prepared by a modified version of the Perkin condensation
[3] and esterification. For the synthesis of the acid 2,2,2-
trifluoroacetophenone (Aldrich) and phenylacetic acid
(Aldrich) were applied, esterification was done in diethyl
ether with diazomethane as methylating agent at room
temperature. Both compounds are new we intend to
elaborate on their synthesis and give characterising NMR,
MS and other physical data in a follow-up paper [4]. The
modified Perkin reaction provided with the E isomer
exclusively.
The molecules are depicted in Fig. 1.
The FT-IR spectra were taken with a BIORAD
FTS-65A/896 spectrometer equipped with DTGS detector.
For measurements in solution CCl4 was used as solvent and
the concentrations of the acid or the ester were set to 10K2
,
10K3 or 10K4 mol/dm3. The KBr pellet technique was
applied for measurements in the solid state (1.2 mg of the
compound in 200 mg KBr). Resolution was 4 cmK1 and 256
scans were collected for a spectrum. Spectra were evaluated
by the WIN-IR software package.
* Corresponding author. Tel.: C36 62 544 288; fax: C36 62 544 200.
´ ´
E-mail address: palinko@chem.u-szeged.hu (I. Palinko).
0022-2860/$ - see front matter q 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.molstruc.2004.11.034

208
J.T. Kiss et al. / Journal of Molecular Structure 744–747 (2005) 207–210
various concentrations are given separately. To make
visualisation easier models for viable C–H/F hydrogen
bonding interactions are also added. The models are of
qualitative nature; molecular modeling by ab initio
methods, suitable for quantitative evaluation is currently
not feasible by our means for such large systems. Moreover,
the set of four molecules may be considered only a very
rough approximation of the solid crystalline materials.
Fig. 1. The molecules studied.
3.1. Intermolecular hydrogen bonding interactions
for the ester
FT-IR spectra were taken for both compounds in the
solid state as well as in solution. The solid-state spectra and
those registered in solution were then compared. When
shifts in the position of certain bands occurred in the solid
state, they were taken as signs of newly arisen interactions,
presumably hydrogen bonds.
Experimental findings concerning this compound in the
solid state and in solutions of various ester concentrations
are summarized in Table 1. Here, those vibrations are listed,
which were shifted on aggregation.
As an example the spectra of the crystalline sample and a
solution of 10K3 mol/dm3 are shown in Fig. 2.
For visualisation of the possible C–H/F close contacts
four of the ester and acid (dimer of the dimer acids bonded via
hydrogen bonding) molecules, respectively, were optimised
to gradient norm 0.1 by the AM1 [5] and the PM3 [6] routines
included in the HyperChem package [7].
From the data of Table 1 two observations may be
deduced. First, there is no aggregation through intermole-
cular hydrogen bonds in solution (there is no shift in the
vibrations on dilution). Second, there are intermolecular
hydrogen bonds in the crystalline ester and in these close
contacts the carbonyl oxygen (1725 cmK1), the fluorine atom
of the CF3 group (1120 cmK1), the aromatic and the methyl
hydrogens (vibrations around 3000 cmK1) are involved.
(It is to be noted that a band separately appearing at
3. Results and discussion
The experimental findings obtained by FT-IR spec-
troscopy on the solid compounds as well as the solution of
Table 1
Those infrared frequencies for the crystalline and the dissolved ester, which were shifted on dissolution
c (mol/dm3)
Position of vibrations (n*) (cmK1
)
Solid
10K2
10K3
10K4
Da
3084
3086
3086
3086
2
3023
3027
3027
3027
4
2958
2952
2952
2952
K6
1725
1739
1739
1739
14
1120
1130
1130
1130
10
1016
1020
1020
1020
4
962
967
967
967
5
a
DZn*liquidKn*solid
.
Fig. 2. The FT-IR spectra of the ester: (a) in the solid state, and (b) in CCl4 solution of 10K3 mol/dm3 concentration (bands shifted on dissolution are marked by
arrows).

J.T. Kiss et al. / Journal of Molecular Structure 744–747 (2005) 207–210
209
and the PM3 methods (the latter is given in parentheses)
indicating hydrogen bonding.
3.2. Intermolecular hydrogen bonding interactions
for the acid
Experimental findings concerning this molecule are
summarised in Table 2. Here, those vibrations are listed
which were shifted on aggregation.
As an example the spectra of the crystalline sample and a
solution of 10K3 mol/dm3 are shown in Fig. 4.
Fig. 3. The tetramer of the ester kept together by (aromatic)C–H/F and
(methyl)C–H/F hydrogen bonds (bond length data are from the AM1
model, data from the PM3 model are in parentheses).
Data in Table 2 reveal that aggregation of the dimers
only occurs in the crystalline acid, since the position of
the bands shifted on going from the solid state to
solution did not move on dilution. The downward shifts
of C–F stretching vibration (from 1274 cmK1 in solution
to 1264 cmK1 for the crystalline acid) and those of C–H
stretching and aromatic deformation vibrations indicate
that in the solid state the dimers are kept together by
(aromatic)C–H/F hydrogen bonds. Interestingly, the
CaO stretching vibrations is also downshifted in the
solid state relative to what was measured in solution.
Because of steric reasons there is no chance for the
existence of (aromatic)C–H/(carbonyl)O hydrogen
bonds, however, cyclic oligomers (via (carbonyl)O/HO
interaction) can be formed easily [9]. Further support for
such an oligomerisation is the structured OH band in
3001 cmK1—aromatic C–H stretching—in the solution
spectra is a weak shoulder at 2998 cmK1 in the solid-state
spectrum on the band at 3024 cmK1.) Modeling even by the
semiempirical methods reveal that (aromatic)C–H/(carbo-
nyl)O hydrogen bonds cannot exist because of steric
reasons but (methyl)C–H/(carbonyl)O interaction is feas-
ible. A model displayed here (Fig. 3) shows viable
(aromatic)C–H/F and (methyl)C–H/F hydrogen bonds.
(In the real sample beside the hydrogen attached to the 40
aromatic carbon atom, other hydrogen atoms of the aromatic
ring(s) may also enter to such interactions.)
Distances between the heavy atoms are within the sum of
Bondi’s van der Waals radii (317 pm [8]) by both the AM1
Table 2
Those infrared frequencies for the crystalline and the dissolved acid, which were shifted on dissolution
c (mol/dm3)
Position of vibrations (n*) (cmK1
)
Solid
10K2
10K3
10K4
Da
3084
3086
3086
3086
2
3025
3027
3027
3027
2
1703
1708
1708
1708
5
1264
1274
1274
1274
10
1178
1181
1181
1181
3
900
904
904
904
4
a
DZn*liquidKn*solid
.
Fig. 4. The FT-IR spectra of the acid: (a) in the solid state, and (b) in CCl4 solution of 10K3 mol/dm3 concentration (bands shifted on dissolution marked
by arrows).

210
J.T. Kiss et al. / Journal of Molecular Structure 744–747 (2005) 207–210
and the PM3 methods (the latter is given in parentheses)
indicating hydrogen bonding.
4. Conclusions
Substituting the olefinic hydrogens in 2,3-diphenyl
propenoic acid and its methyl ester by CF3 groups changed
the aggregation properties of both compounds. The
(aromatic)C–H/(carbonyl)O hydrogen bond, the most
common interaction for the unsubstituted ester, disappeared
and the aggregates, only found in the solid state, became the
(methyl)C–H/(carbonyl)O and two types of C–H/F close
contacts (the carbon is either part of the aromatic rings or the
methyl group). For the acid, the interactions responsible for
the short-range order (dimer and cyclic oligomer formation
that is) are the (carbonyl)O/HO hydrogen bonds, while
long-range ordering (only found in the solid state) is due to
(aromatic)C–H/F hydrogen bonds. The (aromatic)C–
H/(carbonyl)O hydrogen bonds, typical in keeping
together a network of unsubstituted acid dimers, do not
exist because of steric reasons.
Fig. 5. The dimers of the acid dimer kept together by (aromatic)C–H/F
hydrogen bonding (bond length data are from the AM1 model, data from
the PM3 model are in parentheses).
the 2970–2560 cmK1 region in the solid state. This part
of the spectrum gradually disappears on dilution. In
solution, part of the dimers also falls apart indicated by
the appearance of the band at 3509 cmK1 (such a band
was not found in the solid-state IR spectrum). The
relative weight of this band increases on dilution. An
exemplary model displayed in Fig. 5 shows the viable
(aromatic)C–H/F hydrogen bond keeping the dimer of
the dimers together in an arrangement where the
hydrogen on the 40 aromatic carbon is in close contact
with a fluorine atom of the CF3 group. (In the real
sample other hydrogen atoms of the aromatic ring(s) may
also enter to such interactions.)
References
´
[1] A. Kukovecz, I. Palinko, J. Mol. Struct. 482/483 (1999) 463.
´
´
´ ´ ´
[2] J.T. Kiss, K. Felfoldi, T. Kortvelyesi, I. Palinko, Vib. Spectrosc. 22
¨
¨
(2000) 63.
¨
´ ´
[3] K. Felfoldi, M. Sutyinszky, N. Nagy, I. Palinko, Synth. Commun. 30
(2000) 1543.
´
´
´
¨
[4] K. Felfoldi, Sz. Cserenyi, I. Palinko in preparation.
[5] M.J.S. Dewar, E.G. Zoebisch, E.F. Healy, J.J.P. Stewart, J. Am. Chem.
Soc. 107 (1985) 3902.
[6] J.J.P. Stewart, J. Comput. Chem. 10 (1989) 209 (see also p. 221).
[7] HyperChem 7.0, Hypercube, Inc., Gainesville, FL, USA, 2001.
[8] A. Bondi, J. Phys. Chem. 68 (1964) 441.
Distances between the heavy atoms are within the sum of
Bondi’s van der Waals radii (317 pm [7]) by both the AM1
´ ´
[9] I. Palinko, J.T. Kiss, Mikrochim. Acta 14 (1997) 253.