Comparative study on the dynamics, polarity, and thermal properties of the isomers of (4-acetyloxyphenyl)-(chlorophenyl)-methanone

Article abstract of DOI:10.1063/1.472681

The effect of the location of the chlorine atoms on the dynamics, polarity, and thermal properties of the isomers derived from (4-acetyloxyphenyl)-(chlorophenyl)-methanone is discussed. The changes occurring in the dipole moments along the trajectories in

Full text of DOI:10.1063/1.472681

Comparative study on the dynamics, polarity, and thermal properties of the isomers of
(4acetyloxyphenyl)(chlorophenyl)methanone
Enrique Saiz, Cristina Alvarez, Evaristo Riande, Mauricio R. Pinto, and Catalina Salom
Citation: The Journal of Chemical Physics 105, 8266 (1996); doi: 10.1063/1.472681
View online: http://dx.doi.org/10.1063/1.472681
View Table of Contents: http://scitation.aip.org/content/aip/journal/jcp/105/18?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Correlation of Kauzmann temperature with odd–even effect in nalkanes
J. Chem. Phys. 105, 11364 (1996); 10.1063/1.472987
Thermodynamic length, time, speed, and optimum path to minimize entropy production
J. Chem. Phys. 105, 11220 (1996); 10.1063/1.472897
A thermodynamic distance criterion of optimality for the calculation of free energy changes from computer
simulations
J. Chem. Phys. 105, 10072 (1996); 10.1063/1.472836
Comment on ‘‘The effect of charged impurities on a glass transition in a polar medium’’ [J. Chem. Phys. 104, 664
(1996)]
J. Chem. Phys. 105, 8979 (1996); 10.1063/1.472734
Response to ‘‘Comment on ‘The effect of charged impurities on a glass transition in a polar medium’’’ [J. Chem.
Phys. 105, 8979 (1996)]
J. Chem. Phys. 105, 8981 (1996); 10.1063/1.472629
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.16.124 On: Mon, 24 Nov 2014 22:14:55

Comparative study on the dynamics, polarity, and thermal properties
of the isomers of (4-acetyloxyphenyl)-(chlorophenyl)-methanone
Enrique Saiz
´
´
´
´
Departamento de Quımica-Fısica, Universidad de Alcala, 28871-Alcala de Henares, Spain
Cristina Alvarez and Evaristo Riande
´
´
Instituto de Ciencia y Tecnologıa de Polımeros (CSIC), 28006 Madrid, Spain
Mauricio R. Pinto
´
Instituto de Macromoleculas Eloisa Mano, UFRJ, Rio de Janeiro, Brazil
Catalina Salom
Departamento de Materiales, ETSIA, UPM, 28037-Madrid, Spain
͑Received 27 June 1996; accepted 6 August 1996͒
The effect of the location of the chlorine atoms on the dynamics, polarity, and thermal properties of
the isomers derived from ͑4-acetyloxyphenyl͒-͑chlorophenyl͒-methanone is discussed. The changes
occurring in the dipole moments along the trajectories in their respective conformational spaces are
rather large ͑between 6.0 and 0.3 D͒ for ͑4-acetyloxyphenyl͒-͑2-chlorophenyl͒-methanone ͑2CPM͒,
slightly lower for ͑4-acetyloxyphenyl͒-͑3-chlorophenyl͒-methanone ͑3CPM͒, and relatively small
͑between 2 and 4 D͒ for ͑4-acetyloxyphenyl͒-͑4-chlorophenyl͒-methanone ͑4CPM͒. The values of
the mean-square dipole moment of 2CPM, 3CPM and 4CPM, calculated from the evolution of their
dipole moments along their respective trajectories, are 15.9, 13.1, and 11.2 D2, respectively, in very
good agreement with the experimental results 15.8, 13.4, and 11.2 D2, respectively. Both melting
and glass transition temperatures of the compounds are discussed in terms of their respective
conformational entropies. © 1996 American Institute of Physics. ͓S0021-9606͑96͒51442-X͔
INTRODUCTION
relaxation time anomalously increases and its temperature
dependence is governed by the free volume; in other words,
the relaxation time is related to the temperature by the
Vogel–Tammann–Fulcher ͑VTF͒ equation.3–5 A careful ex-
amination of the relaxation processes shows a single maxi-
mum frequency at TӷTg . However, as the temperature ap-
proaches Tg , the peak splits into a pair of maxima, the slow
␣ peak, or glass–liquid relaxation, and the faster ␤ peak, or
secondary process. This behavior can be explained by as-
suming that interactions between atoms and molecules are
comprised in a potential energy function V͑r1,r2,...,rN͒
which depends on the spatial location ri for each of these
atoms, molecules, etc. As the temperature of the liquid de-
creases, the configuration point of the system, R(t), is forced
into regions of increasingly rugged and heterogeneous topog-
raphy in order to seek the ever deeper basins. As the tem-
perature comes closer to Tg , the rarer and more widely sepa-
rated these deep basins must be.6,7 The ␣ or glass–liquid
relaxation entails long-range transitions between deep basins
whereas the ␤ relaxation is associated with elementary tran-
sitions between neighboring basins separated by low barrier
energies.
The chemical structure governs the topology of the con-
figurational space and, consequently, the dynamics of liq-
uids. Among liquids made up by homologous flexible mo-
lecular chains, the larger is their conformational entropy, the
lower is their glass transition temperature.8,9 However, the
height of barrier energies separating conformational states
seems to be mainly responsible for the magnitude of Tg . In
general, the presence of rigid residues in molecular chains
will hinder conformational transitions about the skeletal
The return to equilibrium of liquids perturbed by an ex-
ternal force field, as monitored by a physical property P
sensitive to the perturbation, is customarily expressed by the
simple exponential P(t)ϪPeϭ(PoϪPe)exp͑Ϫt/␶͒, where ␶
and Pe are, respectively, the relaxation time associated with
the relaxation process and the value of the property P at
equilibrium. However, it is also well known that most of the
relaxation processes are not described by a single relaxation
time but rather by a distribution of relaxation times whose
breadth increases as the complexity of the system increases.
For some liquids, such as polymers, relaxation times may
spread from picoseconds to seconds or larger, depending on
temperature, concentration and molecular weight.1,2
When a liquid is cooled, a temperature is reached below
which the ability of the liquid to flow is lost, becoming a
glass. The temperature at which this occurs, called glass tran-
sition temperature, depends on the cooling rate. Glass–liquid
transition is a general phenomenon exhibited by uncrystalli-
zable systems, but even those with capability to develop
crystallinity, when properly supercooled, also can become
structurally a glass. In the glassy state there are noncrystal-
line packing modes for the atoms and molecules which,
though of higher energy than those of the crystalline state,
still have sufficiently low energy for the molecules can easily
assemble. So long as the relaxation times are significantly
shorter than the time scale of the experiment, the liquid
moves in the configurational space among basins of similar
depth. However, as the temperature of the liquid declines
coming near the glass transition temperature, the mean-
8266
J. Chem. Phys. 105 (18), 8 November 1996
0021-9606/96/105(18)/8266/8/$10.00
© 1996 American Institute of Physics
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.16.124 On: Mon, 24 Nov 2014 22:14:55

Saiz et al.: Isomers of (4-acetyloxyphenyl)-(chlorophenyl)-methanone
8267
bonds and the glass-transition temperature will be shifted to
higher values. The same will occur with the location of the ␤
relaxation process in the loss-temperature plane.
In order to get a deeper insight into the dynamics of
liquids, it is important to obtain information on the effect of
the chemical structure on the evolution with time of the
physical properties of a series of homologous molecules. In
this work, the time evolution of the dipole moment
of the asymmetric chloro-isomers of ͑4-acetyloxyphenyl͒-
g ͑0.16 mol͒ of 3-chlorobenzoic acid and 41.6 g ͑0.35 mol͒
of thionyl chloride in 50 mL of toluene was refluxed for 4 h.
The resulting homogeneous solution was vacuum distilled to
remove the solvent and the excess of thionyl chloride. The
remaining oil was diluted with 25 mL of carbon disulfide and
the resulting solution was added to a mixture of 13.5 g
͑0.047 mol͒ of triphenyl borate and 56.0 g ͑0.42 mol͒ of
aluminum chloride in 25 mL of carbon disulfide. The reac-
tion mixture was refluxed for 6 h, poured into an equal mix-
ture of ice and 6N HCl and then heated until complete neu-
tralization of the catalyst. A clear brown waxy residue was
separated and dissolved in 300 mL of 1N NaOH to promote
the displacement of the protecting borate group. The result-
ing solution was treated with activated charcoal, filtered and
acidulated with concentrated HCl. The solid product was
crystallized from methanol/water ͑60/40͒. Yield: 10.4 g
͑27.9%͒.
͑chlorophenyl͒-methanone,
specifically,
͑4-acetyl-
oxyphenyl͒-͑2-chlorophenyl͒-methanone ͑2CPM͒ and ͑4-
acetyloxyphenyl͒-͑3-chlorophenyl͒-methanone ͑3CPM͒ is
followed by molecular dynamics ͑MD͒, and further com-
pared with that of the symmetric isomer, ͑4-
acetyloxyphenyl͒-͑4-chlorophenyl͒-methanone ͑4CPM͒.10 It
is expected that the trajectories of the isomers in the confor-
mational space will be strongly influenced by interactions
between the ketone group and the chlorine atoms. The reli-
ability of the MD calculations is checked by comparing the
values of the mean-square dipole moment calculated from
the conformational trajectories of the molecules with those
experimentally determined. Attention is also paid to relate
the molecular flexibility, as expressed by the conformational
entropy, to the development of crystallinity in these com-
pounds. This work forms part of a more general project car-
ried out in our laboratories dealing with the study of the
influence of the structure of the side groups on the relaxation
processes of acrylic polymers.
(4-acetyloxyphenyl)-(3-chlorophenyl)-methanone
(3CPM) (4) and (4-acetyloxyphenyl)-(2-chlorophenyl)-
methanone (2CPM) (5)
Both compounds were produced by reacting their re-
spective phenols ͑2͒ and ͑3͒ with acetic anhydride in an al-
kaline medium. The acetylated compounds were purified by
passing through a chromatographic column using a basic alu-
minum oxide bead and dichloromethane as eluent; finally ͑4͒
was crystallized from toluene/hexane ͑1:1͒ whereas ͑5͒,
which is liquid at room temperature, was distilled at high
vacuum. The average yield was 80%.
13C-NMR ͑CDCl3͒ for ͑4͒: ␦ 21.5 ͑CH3–͒; 127.8, 129.0,
129.4, 129.6, 130.1, 131.9 ͑Car–H͒; 134.1, 134.3, 139.5,
143.6 ͑Car–͒; 168.6 ͑–COO–͒; 194.7 ͑–CO–͒. 13C-NMR
͑CDCl3͒ for ͑5͒ ␦ 21.0 ͑CH3–͒; 121.7, 126.6, 128.9, 129.9,
131.1, 131.5 ͑Car–H͒; 131.0, 133.8, 138.2, 154.7 ͑Car–͒;
168.5 ͑–COO–͒; 193.8 ͑–CO–͒.
EXPERIMENT
Materials
Aluminum chloride ͑Merck͒, boric acid ͑Riedel͒,
3-chlorobenzoic acid and 2-chlorobenzoic acid ͑Aldrich͒
were used as received. Thionyl chloride ͑Fluka͒, phenol
͑Merck͒ and acetic anhydride ͑Merck͒ were distilled prior to
use. Carbon disulfide ͑Merck͒ was dried with MDI and dis-
tilled. All other analytical grade solvents and other chemicals
from commercial sources were used as received.
Thermal characterization of the compounds
The synthesis of ͑4-acetyloxyphenyl͒-͑4-chlorophenyl͒-
methanone used in the thermal experiments is described
elsewhere.10 Values of both the glass transition and melting
temperature of the three isomers were obtained with a DSC4
Perkin–Elmer calorimeter at a heating rate of 10°/min. The
thermogram corresponding to 2CPM exhibits a single endot-
herm in the neighborhood of Ϫ26 °C corresponding to the
glass–liquid transition. The thermogram of 3CPM presents
an ostensible peak corresponding to a melting process which
seems to reflect some kind of polyformism in the crystal.
However, the thermogram of quenched 3CPM from the melt
exhibits a well developed glass–liquid transition, located in
the vicinity of Ϫ47 °C, followed by an exothermic peak cor-
responding to the crystallization of liquid 3CPM and, finally,
an endotherm arising from the melting process of crystalline
3CPM. This behavior is reflected in Fig. 1. The thermogram
for 4CPM only presents an endotherm associated with the
crystalline→liquid transition. Quenching of molten 4CPM
does not produce a supercooled liquid, as occurs with 3CPM,
but a crystalline solid.
Triphenyl borate(1)
A mixture of 300.0 g ͑3.18 mol͒ of phenol and 31.0 g
͑0.5 mol͒ of boric acid was slowly distilled at normal pres-
sure until the temperature at the top of the fractionation col-
umn reached 180 °C. The excess of phenol was removed by
vacuum distillation. The solid residue was column chromato-
graphed using an activated aluminum oxide bead and CH2Cl2
as eluent. The eluent solution was evaporated and the residue
of ͑1͒ was stored under vacuum to keep it dry. Yield 82.7 g
͑57%͒.
(4-hydroxyphenyl)-(3-chlorophenyl)-methanone (2)
and (4-hydroxyphenyl)-(2-chlorophenyl)-methanone
(3)
A general route was employed to produce both com-
pounds, and that corresponding to ͑4-hydroxyphenyl͒-͑3-
chlorophenyl͒-methanone is as follows. A suspension of 25.0
J. Chem. Phys., Vol. 105, No. 18, 8 November 1996
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.16.124 On: Mon, 24 Nov 2014 22:14:55

8268
Saiz et al.: Isomers of (4-acetyloxyphenyl)-(chlorophenyl)-methanone
FIG. 1. Thermogram for quenched ͑4-acetyloxyphenyl͒-͑3-chlorophenyl͒-
methanone ͑3CPM͒, showing from low to high temperature the glass–
liquid, crystallization and melting processes. The insert shows the glass–
liquid transition in more detail.
FIG. 2. Schematical representation of the planar all-trans conformation
for ͑4-acetyloxyphenyl͒-͑2-chlorophenyl͒-methanone ͑2CPM͒ and ͑4-
acetyloxyphenyl͒-͑3-chlorophenyl͒-methanone ͑3CPM͒ in which the value
␾iϭ180° was assigned to all the rotational angles. Arrows pointing from
negative to positive centers of charges indicate the approximate direction of
the contributions to the total dipole moment of each molecule.
Mean-square dipole moments
Experimental values of the mean square dipole moment
͗␮ ͘ of 3CPM and 2CPM were obtained in benzene solutions
2
using the procedure outlined by Guggenheim and Smith.11,12
The method implies the determination of the total polariza-
tion of the solute which is proportional to the experimental
value of d⑀/dw in the limit w→0, where ⑀ and w are, re-
spectively, the dielectric permittivity of the solution and the
weight fraction of solute. The polarization induced by the
distortion of the electronic clouds by the electric field, which
is proportional to dn/dw, where n is the index of refraction
of the solution, was measured and subtracted from the total
polarization. The atomic polarization was considered to be
and C–Car are represented by ␾1, ␾2, and ␾ respectively.
3
Values of 180° were assigned to the trans states of these
angles ͑i.e., as they are drawn in Fig. 2͒.
Molecular dynamics ͑MD͒ simulations for the isolated
molecules in the gaseous state were carried out with the
SYBYL modeling package13 employing the Tripos force
field.14 A value of ⑀ϭ4 was assigned to the dielectric permit-
tivity for the evaluation of the Coulombic contributions to
the potential energy. The geometry of each molecule was
first optimized with respect to all bond lengths, bond angles,
and rotations, and the resulting conformation was used as
starting point for the MD trajectories in which the Verlet
algorithm15 was employed to integrate the Newton equation
of motion for each atom with a given value of time step ␦,
that was set to 1 fs ͑10Ϫ15 s͒ in all our calculations. The
molecules were first warmed up from 0 K to the working
temperature with increments of 20 K and allowing a relax-
ation interval of 500 time steps in each intermediate tempera-
ture. Once the working temperature was reached, the simu-
lations were continued for 4 106 time steps ͑i.e., during 4 ns͒
at constant temperature and the data of interest were re-
corded every 500 fs.
2
negligible. The values at 30 °C of d⑀/dw, dn/dw, and ͗␮ ͘
for 3CPM and 2CPM are given in Table I. In this table the
values of these quantities for ͑4-acetyloxyphenyl͒-͑4-
chlorophenyl͒-methanone ͑4CPM͒, taken from Ref. 10, are
also shown.
MOLECULAR DYNAMICS CALCULATIONS
The planar all-trans structures of both 2CPM and 3CPM
are represented in Fig. 2. Rotational angles O–Car, Car–C,
TABLE I. Summary of dielectric results at 30 °C for ͑4-acetyloxyphenyl͒-
͑2-chlorophenyl͒-methanone
͑2CPM͒,
͑4-acetyloxyphenyl͒-͑3-chloro-
phenyl͒-methanone
methanone ͑4CPM͒.
͑3CPM͒,
͑4-acetyloxyphenyl͒-͑4-chlorophenyl͒-
Partial charges, computed with the MOPAC program and
the AM116,17 procedure, were assigned to each atom of the
molecules. The values thus obtained are shown in Fig. 3.
These charges were used both to calculate Coulombic inter-
actions and to evaluate the dipole moment of each one of the
conformations analyzed along the MD trajectories.
2
Compound
d⑀/dw
2n1dn/dw
͗␮ ͘,D2
2CPM
3CPM
4CPMa
6.33
5.38
4.46
0.25
0.19
0.15
15.8
13.4
11.2
The a priori probability distribution for the rotational
angles ␾1, ␾2, and ␾3, taken as independent of their neigh-
aTaken from Ref. 10.
J. Chem. Phys., Vol. 105, No. 18, 8 November 1996
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.16.124 On: Mon, 24 Nov 2014 22:14:55

Saiz et al.: Isomers of (4-acetyloxyphenyl)-(chlorophenyl)-methanone
8269
FIG. 5. Distribution of probabilities for the ␾2 and ␾3 rotational angles ͑i.e.,
ar ar
ar
*
rotations around C C –C C bonds in the ketone group of ͑4-
acetyloxyphenyl͒-͑3-chlorophenyl͒-methanone ͑3CPM͒. The asterisks repre-
sent the actual results of MD simulations while the solid line is a least
squares fitting drawn to show up the shape of the maxima.
FIG. 3. Partial charges ͑in electronic units of charge͒ assigned to each atom
of ͑4-acetyloxyphenyl͒-͑2-chlorophenyl͒-methanone ͑2CPM͒ and ͑4-
acetyloxyphenyl͒-͑3-chlorophenyl͒-methanone ͑3CPM͒. These charges were
used for all the calculations presented in this work. See the text for details.
4 for ␾1, and Fig. 5 for ␾2 and ␾3. The rotational behavior of
both 2CPM and 3CPM is practically identical and coincides
also with the results reported for 4CPM in a previous work.10
Thus the preferred orientations for the ester residue are those
placing ester and phenyl groups roughly perpendicular to
each other ͑i.e., ␾1ϭϮ90°͒. On the contrary, the ketone resi-
due is preferently located in the same plane that the phenyl
ring ͑i.e., ␾2,␾3ϭ0°, 180°͒.
bors, was computed by evaluating the fraction of conforma-
tions along the MD trajectories in which the value of one of
these angles lies within an interval of 10° ͑for instance, the
fraction indicated below for ␾1ϭ40° was computed by
counting the number of conformations in which
35°Ͻ␾1Ͻ45°͒. The results at 300 K are summarized in Fig.
The distribution of conditional probabilities for the si-
multaneous rotations of ␾2,␾ is also quite similar for the
3
three isomers. Figure 6 shows, as an example, the 2D and 3D
representations of this distribution for 3CPM at 300 K. The
most important feature of this distribution is that the pre-
ferred states are not placed at the combinations of ␾2,␾3ϭ0°,
180° ͑which are the probability maxima for these angles
when taken as independent͒, but displaced in ca. 30° from
those states with displacements of the same sign in both
angles. A more quantitative information about these prob-
abilities is summarized in Table II where the sum of prob-
abilities computed in the vicinities of each maxima is col-
lected. The differences among these states are rather small
and probably they are within the limit of accuracy of the
calculations.
The evolution of the dipole moment with time at 300 K,
represented in Fig. 7, shows a rather fast conformational in-
terconversion from high to low polarity states. Values of
mean-square dipole moments obtained along the MD trajec-
tories are in excellent agreement with the experimental re-
sults.
FIG. 4. Distribution of probabilities for ␾ rotational angles ͑i.e., rotations
1
around the O–Car bond of the ester residue͒. The curves are similar for the
three isomers. The asterisks represent the actual results of MD simulations
while the solid line is a least squares fitting drawn to show up the shape of
the maxima.
DISCUSSION
Although the actual calculations of dipole moments
along the MD trajectories were performed with partial
J. Chem. Phys., Vol. 105, No. 18, 8 November 1996
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.16.124 On: Mon, 24 Nov 2014 22:14:55

8270
Saiz et al.: Isomers of (4-acetyloxyphenyl)-(chlorophenyl)-methanone
TABLE II. Conditional probabilities for the ␾2 , ␾ pair of rotations com-
3
puted within Ϯ40° intervals around the maxima of ͑4-acetyloxyphenyl͒-͑2-
chlorophenyl͒-methanone ͑2CPM͒ and ͑4-acetyloxyphenyl͒-͑3-chloro-
phenyl͒-methanone ͑3CPM͒.
2CPM isomer
␾
3
␾
Ϫ30
30
150
210
2
Ϫ30
30
150
210
0.105
0.088
0.097
0.098
0.077
0.089
0.117
0.127
3CPM isomer
␾
3
␾
Ϫ30
30
150
210
2
Ϫ30
30
150
210
0.084
0.097
0.099
0.084
0.098
0.086
0.096
0.101
actual MD calculations. If these three group contributions
were exactly parallel, the molecule would have a maximum
dipole moment of 6.3 D, while if they were antiparallel, the
resulting dipole moment would be ca. 0.3 D. An inspection
of Fig. 7 reveals that these are approximately20 the limits
between which the actual dipole moments of 2CPM and
3CPM oscillate. In fact the oscillations are slightly larger in
2CPM than in 3CPM because in the former case the relative
orientation of ␮ and ␮ changes more with the ␾ rotation
2
3
3
than in the later isomer. Assuming that there was no corre-
lation between the three contributions,
a value of
͗␮ ͘ϭ⌺␮2i ϭ14.5 D2 would be obtained for both molecules.
This value lies between the experimental results for these
two molecules, which indicates that positive correlations are
slightly favored in the case of 2CPM, while the situation is
just the opposite in the 3CPM isomer.
2
FIG. 6. Distribution of probabilities, in 2D and 3D representations, for the
ar ar
ar ar
*
simultaneous rotations around the pair of bonds C C –C –C C in the
ketone residue ͑␾ and ␾ rotational angles͒ of ͑4-acetyloxyphenyl͒-͑3-
chlorophenyl͒-methanone ͑3CPM͒.
2
3
Addition of the three contributions for each one of the
preferred conformations ͑i.e., the rotational isomers͒ summa-
rized in Table IV gives the dipole moments collected in
Table III. Mean square dipoles obtained with these values,
assuming the same weight for all isomers, are summarized in
the last row of Table III. Weighting the rotational isomers
with the probabilities indicated in Table II produces rather
small variations of these results. The values obtained with
this latter procedure overestimate the experimental results,
although the trend ͗␮ ͘ ͑2CPM͒Ͼ͗␮ ͘ ͑3CPM͒Ͼ͗␮ ͘
͑4CPM͒ coincides with that experimentally found. Conse-
quently, calculations of these dipole moments with the stan-
dard rotational isomeric states ͑RIS͒ scheme would require
to use smaller dipole contributions of each group than those
indicated above.
charges, it is possible to present an approximate and more
intuitive analysis by adding group contributions to the dipole
moment. The approximate directions of these contributions
coming from ester, ketone and chlorophenyl groups, are rep-
resented by means of arrows ͑pointing from negative to posi-
tive centers of charges͒ in Fig. 2. Values of these contribu-
tions can be taken from experimental values of model
compounds reported in the literature.18 Thus a value of
␮1ϭ1.7 D ͑experimental result obtained for phenyl acetate͒
2
2
2
with a direction defined by an angle ␶ϭ123° between the
19
*
dipole vector and the CH3–C bond was used as contribu-
tion for the ester group; a dipole ␮2ϭ3.0 D ͑average of ex-
perimental results obtained for benzophenone͒ lying along
the CvO bond was used for the ketone residue, and ␮3ϭ1.6
D ͑from chlorobenzene͒ in the direction of the Car–Cl bond
was assigned to the chlorophenyl group. The results of this
approximate analysis are summarized in Table III along with
the experimental values and the results obtained with the
The location of the chlorine atom in ͑4-acetyl-
oxyphenyl͒-͑chlorophenyl͒-methanone isomers, strongly af-
fects the thermal properties of the corresponding isomer.
Thus whereas 3CPM and 4CPM are crystalline compounds
with melting points of 89 and 126 °C, respectively, 2CPM is
liquid at room temperature with a glass transition tempera-
J. Chem. Phys., Vol. 105, No. 18, 8 November 1996
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.16.124 On: Mon, 24 Nov 2014 22:14:55

Saiz et al.: Isomers of (4-acetyloxyphenyl)-(chlorophenyl)-methanone
8271
TABLE III. Summary of experimental and calculated values at 300 K of
dipole moments for ͑4-acetyloxyphenyl͒-͑2-chlorophenyl͒-methanone
͑2CPM͒, ͑4-acetyloxyphenyl͒-͑3-chlorophenyl͒-methanone ͑3CPM͒ and ͑4-
acetyloxyphenyl͒-͑4-chlorophenyl͒-methanone ͑4CPM͒.
Parameter
2CPM
3CPM
4CPMa
2
͗␮ ͘ Experim.
15.7
15.9
3.80
14.5
19.6
13.4
13.1
3.47
14.5
19.6
11.2
11.2
3.28
14.5
2
͗␮ ͘ MD calc.
͗␮͘ MD calc.
2
͗␮ ͘ Uncorrel.
2
͗␮ ͘ RIS
aTaken from Ref. 10.
The distribution of the probabilities of the rotational
angles about the bonds flanking the ketone group of 2CPM
are shown at two temperatures in Figs. 8 and 9. Two maxima
ar
*
about C –C bonds are observed which correspond to the
rotational angles ͑0°, 180°͒ at which the phenyl linked to the
ester group is coplanar with the ketone group. As the tem-
perature increases, the conformers population in the neigh-
borhood of these two rotational angles slightly decreases.
Calculations of the rotational population about the bond link-
ar
*
ing the ketone group to the 2-chlorophenyl group ͑C –C ͒
show that the maxima are shifted from 0°, 180° to 60°, 200°.
As the temperature increases the population of rotational
states is nearly similar for the rotational angles lying in the
range 60°–200°, in sharp contrast with what occurs for
3CPM and 4CPM where well defined potential wells about
bonds flanking the ketone group occur. Consequently, strong
intramolecular interactions between the chlorine atom and
the ketone group hinder the coplanar conformation of the
–C6H4–CO–C6H4Cl moiety in 2CPM thus impeding the de-
velopment of crystalline order in this compound.
Since the molecular isomers have similar polarity, it is
expected that their glass transition temperature is governed
by the molecular flexibility which is conveniently expressed
in terms of either the rotational partition function, Z, or the
conformational entropy, ⌬Sc , this latter quantity being more
convenient because it does not depend on the reference state.
The conformational entropy is currently expressed by
FIG. 7. Evolution with time of the total dipole moment of the isomers. From
up to down: ͑4-acetyloxyphenyl͒-͑2-chlorophenyl͒-methanone ͑2CPM͒,
͑4-acetyloxyphenyl͒-͑3-chlorophenyl͒-methanone ͑3CPM͒ and ͑4-acetyl-
oxyphenyl͒-͑4-chlorophenyl͒-methanone ͑4CPM͒.
⌬ScϭϪkB
p ln p ,
i i
͚
i
where kB is the Boltzmann constant and pi is the probability
of the conformation i which also can be written as
ture of Ϫ26 °C. On the other hand, 3CPM, unlike 4CPM,
becomes a supercooled liquid when quenched from the melt,
with a glass transition temperature of ca. Ϫ47 °C. Owing to
the small variations in the polarity of the chloroisomers of
͑4-acetyloxyphenyl͒-͑chlorophenyl͒-methanone, it is ex-
pected that intramolecular interactions rather than intermo-
lecular ones may be held responsible for the differences ob-
served in their thermal properties. Consequently, the fact that
both 4CPM and 3CPM are crystalline at room temperature
whereas 2CPM is liquid suggests that the conformational
space of this latter compound must differ significantly from
that of the crystalline compounds.
exp ϪE /k T͒
͑
i
B
piϭ
,
Z
where Ei is the energy of the conformation i. The fact that
the value of the glass transition temperature of 3CPM,
Ϫ47 °C, is nearly 21 °C below that of 2CPM suggests a
slightly higher freedom of rotation about the two consecutive
Car–CO–Car bonds in the former compound. This fact, rather
than differences in conformational entropies may be held re-
sponsible for the lower Tg of 2CPM. It should be stressed in
this regard that the values conformational entropies associ-
J. Chem. Phys., Vol. 105, No. 18, 8 November 1996
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.16.124 On: Mon, 24 Nov 2014 22:14:55

8272
Saiz et al.: Isomers of (4-acetyloxyphenyl)-(chlorophenyl)-methanone
TABLE IV. Dipole moments (D) calculated by adding group dipoles ͑see Fig. 2͒, for the preferred orientations of ͑4-acetyloxyphenyl͒-͑2-chlorophenyl͒-
methanone ͑2CPM͒ and ͑4-acetyloxyphenyl͒-͑3-chlorophenyl͒-methanone ͑3CPM͒.
2CPM ͑␾1ϭ90͒
2CPM ͑␾1ϭ270͒
␾
␾
3
3
␾
Ϫ30
30
150
210
Ϫ30
30
150
210
2
Ϫ30
30
150
210
2.92
4.51
6.04
3.66
6.04
4.43
3.59
3.03
5.98
4.55
3.02
3.60
4.48
5.98
3.65
2.89
3CPM ͑␾1ϭ90͒
3CPM ͑␾1ϭ270͒
␾
␾
3
3
␾
Ϫ30
30
150
210
Ϫ30
30
150
210
2
Ϫ30
30
150
210
2.69
4.36
2.77
5.43
2.76
2.74
5.42
4.38
2.72
2.73
4.31
5.40
2.79
5.45
2.66
4.28
ated with the pair of bonds indicated above are nearly similar
͑13.2 and 13.0 cal molϪ1 KϪ1 for 3CPM and 2CPM, respec-
tively, at 30 °C͒.
kcal molϪ1 higher than that of the latter compound. It should
be pointed out that the values of both ⌬Hm and Tm were
obtained in conditions far from equilibrium and, conse-
quently, the results can be considered only semiquantitative.
The values of the melting entropy thus obtained are 16.5 and
19.2 cal molϪ1 KϪ1, respectively, for 3PMC and 4PMC. The
melting behavior of these compounds may be critically inter-
preted in terms of the conformational entropy ⌬Sc , which is
related to the entropy of fusion ⌬Sm , as indicated below.
Actually, this parameter can be expressed by the following:
The evaluation of the melting entropy, ⌬Sm , requires to
know with enough precision the melting enthalpy and melt-
ing temperature of the crystalline compounds. However,
whereas the melting peak of 4CPM is sharp and nearly inde-
pendent on the crystallization conditions, the melting endo-
therm of 3CPM shows a strong dependence on the under-
cooling. Thus the value of ⌬Hm for 3CPM increases from
4.11 kcal molϪ1, obtained in the conditions indicated in Fig.
1, to 6.04 kcal molϪ1 if the crystallization of liquid 3CPM is
performed at 30 °C. The melting enthalpy for 4CPM is
somewhat higher than that of 3CPM, its value being 1.65
ץS
ץp
⌬S ϭ͑⌬S ͒ ϩ
dVϭ͑⌬S ͒ ϩ
dV,
͵
͵
ͩ ͪ
ͩ ͪ
m
m
V
m V
ץV
ץT
T
V
where (⌬Sm)V is the melting entropy at constant volume.
Boyd and Starkweather21 have considered that ⌬Sm involves
the conformational entropy ⌬Sc and two additional terms,
ar ar
ar
a
ar ar
*
*
*
*
FIG. 8. Distribution of probabilities about OC C –C ͑O ͒C bonds in
͑4-acetyloxyphenyl͒-͑2-chlorophenyl͒-methanone ͑2CPM͒, at the tempera-
tures indicated.
FIG. 9. Distribution of probabilities about C C ͑O ͒–C C bonds in ͑4-
acetyloxyphenyl͒-͑2-chlorophenyl͒-methanone ͑2CPM͒, at the temperatures
indicated.
J. Chem. Phys., Vol. 105, No. 18, 8 November 1996
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.16.124 On: Mon, 24 Nov 2014 22:14:55

Saiz et al.: Isomers of (4-acetyloxyphenyl)-(chlorophenyl)-methanone
8273
one of them negative, which corresponds to the packing of
the molecules within the crystal, and the other positive, aris-
ing from long distance disorders. Based on metallographic
results, these authors assign the value of 0.86 R to the nega-
tive contribution. According to this
5 G. Tammann and W. Z. Hesse, Anorg. Allgem. Chem. 156, 245 ͑1926͒.
6 C. A. Angell, Science 267, 1924 ͑1995͒.
7 F. H. Stillinger, Science 267, 1935 ͑1995͒.
8 A. E. Tonelli, Macromolecules 7, 632 ͑1974͒; 8, 544 ͑1975͒; 10, 633
͑1977͒.
9
´
L. Garrido, E. Riande, and J. Guzman, Makromol. Chem. Rapid Commun.
4, 725 ͑1983͒.
⌬S ͒ ϭ⌬S Ϫ0.86Rϩ⌬S .
͑
10 R. C. Nunes, M. R. Pinto, E. Saiz, and E. Riande, Macromolecules 28, 211
͑1995͒.
m
V
c
d
It has been argued that for some systems, such as polyethyl-
ene, the molecular packing and long distance disorder con-
tributions to the melting entropy are similar leading to the
assumption ⌬ScХ(⌬Sm)V . The calculated values of ⌬Sc of
crystalline 3CPM and 4CPM compounds at their respective
melting temperatures, 17.1 and 17.3 KϪ1 cal molϪ1 KϪ1, are
close to the semiquantitative values of ⌬Sm indicated above.
The similarity of these values suggest that differences in
packing the molecules in the crystal rather than differences
in the molecular flexibility may be held responsible for the
higher melting temperature of the 4CPM isomer.
11 E. A. Guggenheim, Trans. Faraday Soc. 45, 714 ͑1949͒; 47, 573 ͑1951͒.
12 J. W. Smith, Trans. Faraday Soc. 46, 394 ͑1950͒.
13 Tripos Associates Inc., St. Louis, MO 63144, U.S.A. A complete refer-
ence of the force field is given in Ref. 14.
14 M. Clark, R. D. Cramer III, N. van Opdembosch, J. Comput. Chem. 10,
983 ͑1989͒.
15 M. P. Allen and D. J. Tildesley, Computer Simulation of Liquids ͑Oxford
Science, Oxford, 1992͒.
16 QCPE, Department of Chemistry, Indiana University, Bloomington, IN
47405, USA.
17 The MOPAC program ͑Version 5.0͒ containing the AM1 procedure is in-
cluded in the SYBYL package by agreement between Tripos and QCPE. A
description of the procedure is given in J. J. P. Stewart, J. Comput.-Aid.
Mol. Des. 4, 1 ͑1990͒ and M. J. S. Dewar, E. G. Zoebisch, E. F. Healy,
and J. J. P. Stewart, J. Am. Chem. Soc. 107, 3902 ͑1985͒.
18 A. L. McClellan, Tables of Experimental Dipole Moments ͑Freeman, San
Francisco, 1963͒, Vol. I; Rahara Enterprises, El Cerrito, CA, 1974, Vol.
II; 1989, Vol. III.
ACKNOWLEDGMENT
This work was supported by the DGICYT through Grant
Nos. PB94-0364 and PB92-0773.
19 E. Saiz; J. P. Hummel, P. J. Flory, and M. Plavsic, J. Phys. Chem. 85,
3211 ͑1981͒.
1 W. W. Graessley, Viscoelasticity and Flow in Polymer Melts and Concen-
trated Solutions in Physical Properties of Polymers, edited by J. E. Mark
͑American Chemical Society, Washington, D.C., 1984͒, p. 97.
2 J. D. Ferry, Viscoelastic Properties of Polymers ͑Wiley-Interscience, New
York, 1980͒.
20 The values indicated as low and high limits for the variation of ␮ are only
approximations, since the actual calculation was performed using partial
charges instead of adding the vectors that represent group contributions.
Thus the result depends not only on the charges, but also on the bond
lengths and bond angles, that change along the MD trajectory.
21 H. W. Starkweather and R. H. Boyd, J. Am. Chem. Soc. 64, 410 ͑1960͒.
3 H. Z. Vogel, Phys. 22, 645 ͑1921͒.
4 G. S. Fulcher, J. Am. Ceram. Soc. 8, 339 ͑1925͒.
J. Chem. Phys., Vol. 105, No. 18, 8 November 1996
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
155.33.16.124 On: Mon, 24 Nov 2014 22:14:55