Thermokinetics of the reaction of phenyl glycidyl ether with aniline

Article abstract of DOI:10.1007/s11172-005-0261-0

The thermokinetic curves in the reaction of phenyl glycidyl ether with aniline were calculated for various compositions of the reaction mixture and temperatures. In addition to the main exothermic effect related to the epoxide ring opening, another exothermic effect of unknown nature was observed. The kinetic data obtained are explained in terms of structural changes caused by the self-aggregation of the reaction product molecules. The "kinetic investigation" approach provides a quantitative analysis of calorimetric data.

Full text of DOI:10.1007/s11172-005-0261-0

378
Russian Chemical Bulletin, International Edition, Vol. 54, No. 2, pp. 378—382, February, 2005
Thermokinetics of the reaction of phenyl glycidyl ether with aniline
R. M. Vinnik,a E. A. Miroshnichenko,a and V. A. Roznyatovskyb
ꢀ
aN. N. Semenov Institute of Chemical Physics, Russian Academy of Sciences,
4 ul. Kosygina, 119991 Moscow, Russian Federation.
Fax: +7 (095) 137 8318
bDepartment of Chemistry, M. V. Lomonosov Moscow State University,
1 Leninskie Gory, 119992 Moscow, Russian Federation.
Eꢀmail: vit.rozn@chem.msu.ru
The thermokinetic curves in the reaction of phenyl glycidyl ether with aniline were calcuꢀ
lated for various compositions of the reaction mixture and temperatures. In addition to the
main exothermic effect related to the epoxide ring opening, another exothermic effect of
unknown nature was observed. The kinetic data obtained are explained in terms of structural
changes caused by the selfꢀaggregation of the reaction product molecules. The "kinetic investiꢀ
gation" approach provides a quantitative analysis of calorimetric data.
Key words: "kinetic investigation" approach, thermokinetics, phenyl glycidyl ether, therꢀ
mochemistry, epoxy compounds, selfꢀaggregation.
A mixture of reactants of a known composition was prepared at
~20 °C. A weighed sample of the mixture (0.25 g) was placed in
a standard aluminum cell, and the cell was sealed and placed in
an DSKꢀIII differential scanning calorimeter (Setaram, France).
Kinetic measurements were carried out in the isothermal reꢀ
gime. The reaction product was studied in the dynamic regime,
and the measurements were carried out after the samples were
stored for a long time at ~20 °C. The heating rate with the
temperature scan was 5 °C min–1. The conversion at each time
moment was calculated from the ratio Qt/Qmol, where Qt is the
quantity of heat released at a given moment, and Qmol is the
molar heat of the reaction referred, in all cases, to 1 mole of
functional groups of a deficient reactant.
Epoxide—amine systems are the subject of many
thorough studies, although the kinetics of many reactions
in these systems had not been extensively studied. In parꢀ
ticular, this is true for the reaction of phenyl glycidyl
ether with aniline (Scheme 1).
Scheme 1
Results and Discussion
The typical curve of the rate of heat evolution (Wexp
)
This reaction has a high exothermic effect and, hence,
it can conveniently be studied using the thermokinetic
method.1 However, the association of the reaction prodꢀ
uct molecules and the formation of a new phase result in
the situation when the kinetic equation that takes into
account only the mechanism of reactant consumption
fails to describe the rate of heat evolution (W ) as a funcꢀ
tion of time observed in experiments.
obtained by the kinetic experiment in the isothermic reꢀ
gime is presented in Fig. 1. Some results obtained by
scanning the samples in the dynamic regime are presented
in Fig. 2. The thermograms contain the characteristic
endotherm, whose size depends on the ratio of reactants
in the reaction mixture (see Fig. 2). This peak cannot be
registered upon the repeated scan. An additional heat
release is observed during the kinetic experiment in the
thermograms of the reacting mixtures that contain the
endotherm.
The purpose of this work is to analyze thermochemiꢀ
cal data for the reaction of phenyl glycidyl ether with
aniline by scanning calorimetry.
The thermokinetics of the reaction considered can be
presented by Scheme 2, which shows that the chemical
reaction proceeds via two routes: noncatalytic and autoꢀ
catalytic (their parameters, viz., rate constants (k) and
heats of reactions (Q), are marked by indices 1 and 2,
Experimental
Phenyl glycidyl ether (E) and aniline (A) (both purchased
from Merck) were purified by vacuum distillation under argon.
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 2, pp. 371—375, February, 2005.
1066ꢀ5285/05/5402ꢀ0378 © 2005 Springer Science+Business Media, Inc.

Thermokinetics of phenyl glycidyl ether
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 2, February, 2005
379
W/J L–1 s–1
which can be explained by the formation of aggregated
molecules of the reaction product (Paggr).
Scheme 2 is described by the kinetic equation
W = Qmol,1
kapp,1[E][A] + Qmol,2kapp,2[AE][P – Pinact] +
60
+ Qmol,3kapp,3[(P – Psat) – Pinact][P – Psat],
(1)
Pinact = 0 for P < Psat
,
30
3
1
where kapp,1 and Qmol,1 are the apparent rate constant and
molar heat of the noncatalytic reaction; kapp,2 and Qmol,2
are the apparent rate constant and molar heat of the autoꢀ
catalytic reaction. An additional heat release (Qmol,3) was
observed during the kinetic experiment for all reacted
mixtures that show the endotherms (see Fig. 2). To take
into account the contribution of the additional heat effect
to the exothermic behavior of the reaction, we assumed
that the interaction process involves the period of nucleꢀ
ation and development of crystallization nuclei, and preꢀ
cipitation occurs at a temperature lower than the reaction
temperature. Then in Eq. (1) Pinact is the concentration of
the catalytically active OH groups leaving the reaction
volume during saturation, and Psat is the saturated conꢀ
centration of the crystallization nuclei.
The expression for the autocatalytic reaction was obꢀ
tained under the assumption that the reactants are inꢀ
volved in the equilibrium formation of a 1 : 1 AE comꢀ
plex, which is autocatalytically transformed into the reacꢀ
tion product P with the apparent rate constant kapp,2. This
assumption is based on the kinetic data for the epꢀ
oxide—amine system with different functionalities of the
reacting molecules. For example, we have previously
shown2 that the kinetics of the reaction of diamine with
monofunctional epoxide is based on the autocatalytic reꢀ
action of a complex formed by the reactants. The kinetics
of the epoxide ring opening remains unchanged when
monofunctional epoxide is replaced by its bifunctional
analog.3
2
4
0
50
100
150
200
250 t/min
Fig. 1. Experimental (Wexp) (1, points) and calculated (Wcalc
(1, line) curves of the heat release rate deconvoluted into comꢀ
ponents W1 (2), W2 (3), and W3 (4).
)
Cp/J g–1 °C–1
1.2
0.9
0.6
0.3
a
1
2
40
80
120
T/°C
Cp/J g–1 °C–1
b
0.9
0.6
1
2
0.3
40
80
1
120
T/°C
T/°C
Cp/J g–1 °C–1
c
0.6
2
0.3
40
80
120
Fig. 2. Temperature plots of the heat capacity (Cp) in the reacted
mixture with the ether to aniline ratios: 1.06 (a), 0.7 (b), and
0.47 (c); 1, first scan and 2, second scan (reaction temperaꢀ
ture 100 °C).
In addition to the noncatalytic and autocatalytic reacꢀ
tions, another process contributes to the overall exotherꢀ
mic effect. We interpret this process as the formation of
intermolecular associates by the product molecules folꢀ
lowed by the loss of catalytic activity due to association.
The formation of crystallization nuclei from the reacꢀ
tion product molecules can result in the loss of catalytic
activity. Based on the spectral study of the reaction of
aniline with phenyl glycidyl ether, we assumed that comꢀ
plexes with a hydrogen bond between the OH groups are
formed at high conversions.4 Since these are the OH
groups that are active in the autocatalytic reaction, the
association of the reaction product molecules can result
in the catalytic activity loss. Therefore, the first two terms
of Eq. (1), which describe the consumption of the reacꢀ
tants, can be supplemented by another term that takes
into account structural changes in the reaction mixture
introduced by the molecules formed in the reaction.
respectively). According to the autocatalytic route, the
OH groups that formed catalyze the epoxide ring openꢀ
ing. Index 3 designates an additional heat release process,
Scheme 2
A is aniline, E is phenyl glycidyl ether,
AE is reaction complex,
P is reaction product

380
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 2, February, 2005
Vinnik et al.
Table 1. Thermokinetic data for the reaction of aniline (A) with phenyl glycidyl ether (E) at
70 and 100 °C
70 °C
Qmol,2
100 °C
Qmol,2
[A]/[E] Qmol,1
Qmol,3
Csat
[A]/[E] Qmol,1
Qmol,3
Csat
(%)
(%)
kJ mol–1
kJ mol–1
0.52
0.72
0.86
1.06
1.28
1.54
1.90
3.10
4.50
335.6
169.6
152.8
204.3
204.3
377.2
201.3 107.200
522.5
258.7
89.86
92.95
84.57
85.83
97.97
99.23
14.72
26.75
46.47
38.10
26.13
25.96
0
73.2
69.0
63.5
69.8
74.2
76.3
100
100
100
0.48
0.70
0.90
1.06
2.20
3.06
4.08
029.9
262.0
098.0
068.7
201.0
213.0
201.0
72.5
87.1
91.1
78.2
98.4
15.5
45.7
52.1
25.8
0
63.1
60.6
67.9
73.7
100
100
100
106.00
87.7
0
0
86.08
93.78
0
0
ln(kapp/L mol–1 s–1
)
The Pinact concentration in Eq. (1) is related to the
number of functional groups of the product molecules
leaving the reaction sphere. In this case, Psat is the conꢀ
centration of catalytically active species. Once the value
of Psat is reached an additional heat release commences
accompanied by the establishment of a saturated concenꢀ
tration of the OH groups in the reaction medium; Qmol,3 is
the molar heat of this process; and kapp,3 is the apparent
rate constant of the decrease in the catalytic active species
in the reaction volume.
2
–10
–12
–14
1
The kinetic curve of the heat release calculated by
Eq. (1) and the results of deconvolution of the total rate
of heat release to components W1, W2, and W3 are preꢀ
sented in Fig. 1. Indices 1, 2, and 3 designate the rates of
the noncatalytic and autocatalytic reactions and the selfꢀ
aggregation of the reaction product molecules, respecꢀ
tively. The deviation of the calculated kinetic curve from
the experimental one was estimated by the R factor.5 In
all experiments, the R factor did not exceed 1.5%. The
heats Qmol,1, Qmol,2, and Qmol,3 for the considered proꢀ
cesses are presented in Table 1.
2.7
2.8
2.9
T –1•103/K–1
Fig. 3. Arrhenius dependences of kapp,1 (1) and kapp,2 (2).
obtained data, the consumption of the reactants in the
reaction (see Scheme 1) proceeds mainly via the autoꢀ
catalytic route (see Table 1).
The reaction rate always increases with an increase in
the amount of aniline in the mixtures with an excess of
the latter, while the mixtures with excess epoxide react
more slowly than the stoichiometric mixtures2—7 (see
Table 1). The increase in the conversion rate of the mixꢀ
ture containing excess aniline can be explained by the
catalytic activity of its excessive amount. This conclusion
is confirmed by the data presented in Fig. 4, indicating
that the kapp,2 values are virtually constant in the region
where the reactant ratios change from excess epoxide to
the stoichiometric composition. On further increase in
the aniline content in the reaction mixture a linear deꢀ
pendence of kapp,2 on the excessive (compared to the
stoichiometric composition) concentration of the NH
groups is observed
The initial points in the thermokinetic curve were obꢀ
tained by extrapolation, which especially affects the Qmol,1
values. The rate of the noncatalytic reaction W1 achieves
remarkable values within the first 75 min. The kapp,1 and
kapp,2 values calculated by Eq. (1) satisfactorily fit the
Arrhenius equation with an almost equal slope (Fig. 3).
The apparent activation energy (Ea) values are equal to
50.6 and 51.2 kJ mol–1, respectively. These data agree
with the experimentally found Ea values. For the autoꢀ
catalytic reaction between bifunctional amine and epꢀ
oxide, Ea is 51 kJ mol–1, whereas that for the noncatalytic
reaction is 52 kJ mol–1 6
. The Arrhenius plot for the
noncatalytic reaction has a higher preꢀexponential factor
than that for the autocatalytic reaction.
A=E
kapp,2 = kA(CA – CE ) + kapp,2
,
(2)
0
0
The processes with the heat release rates W1, W2,
and W3 differ by the molar heats Qmol. Due to this, they
were separated quantitatively. As can be seen from the
where the product kA(CA – CE ) is related to the catalytic
is the apparent reacꢀ
tion rate constant at the stoichiometric composition of
0
activity of excess aniline0, kapp,2
A=E

Thermokinetics of phenyl glycidyl ether
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 2, February, 2005
381
kapp,2•104/L mol–1 s–1
the induction period of nucleation and nuclei developꢀ
ment. Analogous twoꢀstep precipitation processes preꢀ
ceded by the formation of a metastable amorphous phase
have been described.9,10
We have previously2 mentioned a tendency to the
structural organization of the molecules produced in the
reaction of phenyl glycidyl ether with mꢀphenylenediꢀ
amine. Unlike the system with aniline analyzed in the
present work, crystallization was observed at the end of
the kinetic experiment. After the reaction completion,
the kinetic curve of heat release was characterized by a
new exotherm and the reacted samples contained a crysꢀ
talline precipitate.
The kinetic model proposed for the description of
the epoxide ring opening by the amino group in the
aniline—phenyl glycidyl ether system reflects structural
changes in the reaction medium in addition to the mechaꢀ
nism of reactant consumption. The exothermic effect obꢀ
served during the reaction is composed of the effects of
three processes, two of which are chemical reactions, and
the third process is an induction period of crystallization
nucleation in the reaction of aniline with phenyl glycidyl
ether. The heat release kinetics of this process calculated
by Eq. 1 as the phase separation process is shown in Fig. 1
(curve 4).
Thus, using the "kinetic investigation" approach we
found that the aniline—phenyl glycidyl ether reaction sysꢀ
tem is a typical example of the structure directing meꢀ
dium in which the main reaction occurs between reacꢀ
tants bound in a stable complex and stable molecular
aggregates are the reaction product.
It has been known long ago that the behavior of the
system can kinetically be studied at the supramolecular
level. Such examples were described for hydrophobic inꢀ
teractions.11,12 However, the appearance of similar strucꢀ
tural aggregates during the reaction in nonaqueous media
is not usually considered, because the selfꢀorganization of
particles is most often ascribed to hydrophobic interacꢀ
tions. The spontaneous aggregation of hexanoic acid in
acetonitrile indicates that a similar mechanism is signifiꢀ
cant for organic media as well.13 Evidently, these are the
solvophobic interactions that contribute mainly to the
stability of molecular aggregates in the reaction of phenyl
glycidyl ether with aniline.
4
2
3
2
1
1
–4
–2
0
2
4
6
[A]0 – [E]0/mol L–1
Fig. 4. Dependences of kapp,2 on excess aniline at 70 (1) and
100 °C (2) ([A]0 and [E]0 are the concentrations of aniline and
phenyl glycidyl ether in the initial mixture, respectively).
the reactants, and CA – CE0 is excess aniline with respect
0
to phenyl glycidyl ether in the starting mixture.
In the region of an excessive concentration of the
epoxide groups, a constant kapp,2 value is inherent in these
systems. The catalytic activity observed for excess amine is
a characteristic feature of the reactions in epoxide—amine
systems and additionally confirms our hypothesis on the
preꢀassociative mechanism of the autocatalytic reaction.3
The process, whose heat release rate is designed
through W3 (see Fig. 1), manifests itself during the reacꢀ
tion as an additional heat release. This process begins
when the concentration of the OH groups in the reaction
medium reaches saturation. The kinetically determined
conversions above which the additional heat release apꢀ
pears (Csat) are presented in Table 1. It follows from these
data that Csat is independent of the experimental condiꢀ
tions and corresponds to conversions of ~70%.
As can be seen from the data in Fig. 2, the thermoꢀ
grams with the endotherms are similar to the melting
thermograms. However, judging from the position of
these peaks observed within 45—80 °C with a minimum
about 70 °C, we can conclude that no crystallization
should occur when the reaction is carried out at temperaꢀ
tures above 80 °C. The kinetic data obtained at 70 and
100 °C are similar, implying that crystallization could
occur only during cooling of the samples, while their
heating during temperature scanning induces precipitate
melting.
Evidently, the intermolecular interaction of the prodꢀ
uct molecules results in their selfꢀaggregation followed by
crystallization. The latter occurs at temperatures lower
that the temperature of a kinetic experiment. Since the
crystallization occurs only in overcooled or oversaturated
solutions (melts), molecular clusters are predominantly
formed at low overcooling or oversaturation.8
Thus, using the reaction of phenyl glycidyl ether with
aniline, we showed that the thermokinetic study of the
process makes it possible to construct the kinetic model of
the reaction involving the kinetic description of both the
epoxide ring opening and supramolecular organization of
the reaction product molecules.
Thus, we explain the additional heat release found
using the "kinetic investigation" approach by the selfꢀ
aggregation of the product molecules preceding crystalliꢀ
zation. This section of the kinetic curve corresponds to
References
1. K. A. Connors, Chemical Kinetics. The Study of Reaction
Rates in Solution, VCH Publ., Inc., New York, 1990, 36.

382
Russ.Chem.Bull., Int.Ed., Vol. 54, No. 2, February, 2005
Vinnik et al.
2. R. M. Vinnik and E. A. Miroshnichenko, Khim. Fiz.,
1986, 5, 1668 [Sov. J. Chem. Phys., 1990, 5, No. 8 (Engl.
Transl.)].
8. J. D. Dunitz, Chem. Commun., 2003, 545.
9. M. L. Chepelevetskii, Zh. Fiz. Khim., 1939, 13, 561 [Russ.
J. Phys. Chem., 1939, 13 (Engl. Transl.)].
3. R. M. Vinnik and V. A. Roznyatovsky, J. Therm. Anal. Cal.,
2003, 73, 819.
10. E. L. Heeley, C. K. Poh, W. Li, A. Maidens, W. Bras, I. P.
Dobnya, A. J. Gleeson, N. J. Terrill, J. A. Fairclough, C. D.
Olmsted, R. I. Ristic, M. J. Hounslow, and A. J. Ryan,
Faraday Discuss., 2002, 122, 343.
11. J. P. Guthrie, Can. J. Chem., 1973, 51, 3494.
12. J. B. Jones and K. D. Gordon, Biochemistry, 1973, 12, 71.
13. S. Moshizuki, Y. Usui, and A. Wakisaka, J. Chem. Soc.,
Faraday Trans., 1998, 94, 547.
4. J. Mijoviс´ and S. Andjelic, Macromolecules, 1995, 28, 2787.
5. B. P. Demidovich and I. A. Maron, Osnovy vychislitel´noi
matematiki [Foundations of Computational Mathematics],
Nauka, Moscow, 1970, 664 pp. (in Russian).
6. R. M. Vinnik and V. A. Roznyatovsky, J. Therm. Anal. Cal.,
2003, 71, 531.
7. R. M. Vinnik, Izv. Akad. Nauk SSSR, Ser. Khim., 1981, 1027
[Bull. Acad. Sci. USSR, Div. Chem. Sci., 1981, 30 (Engl.
Transl.)].
Received July 18, 2003;
in revised form July 12, 2004