Macromolecules 1992, 25, 979-985 979

Mechanistic Modeling of Epoxy-Amine Kinetics. 1. Model

Compound Study
Jovan Mijovic,* Arnon Fishbain,7 and Jony Wijaya*
Department of Chemical Engineering, Polytechnic University, 333 Jay Street, Brooklyn,
New York 11201
Received May 28, 1991; Revised Manuscript Received September 26, 1991

ABSTRACT: An investigation was carried out of the mechanism and rate of epoxy-amine reactions using
model compounds. High-performance liquid chromatography was used to generate experimental results. A
mechanistic reaction model based on the existence of different transition complexes was proposed, and the
corresponding kinetic parameters were evaluated. Excellent agreement was observed between model predictions
and experimental data.

I. Introduction questions from fundamental mechanistic studies. The

See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Since Pierre Castan’s seminal work, over half a century phenomenological approach to modeling epoxy-amine
cure, which has been explored by many authors,7-13
ago,1 epoxy resins have graduated from a laboratory
curiosity to a multitude of applications, becoming an including our group,14-17 is not considered in this study
Downloaded via KYUSHU UNIV on December 14, 2021 at 09:02:04 (UTC).

due to its batch specificity and thus limited applicability

integral part of the adhesives industry, printed board to modeling and optimization of processing from the first
circuitry, and various aircraft and spacecraft designs.
Depending on the intended use, one can nowadays choose
principles.
between numerous epoxy formulations, differing in the The main objectives of the work reported in this paper
are (1) to conduct a fundamental investigation of epoxy-
composition and chemical structure of the ingredients.
Formulations consisting of epoxy resins and aromatic amine reactions using model compounds, (2) to elucidate
amines are the most widely used thermoset polymer the reaction mechanism, and (3) to develop the corre-
matrices in advanced composites, and that fact has sponding mechanistic kinetic model.
provided the major impetus for this study.
Thermoset polymers are processed by cure, which entails II. Experimental Section
conversion of liquid monomers, or prepolymers, into three- A. Materials. The model compound formulation consisted
dimensional networks. The mechanism and kinetics (rate) of aniline (99.55%, Aldrich) and l,2-epoxy-3-phenoxypropane
of cure determine the network morphology, which, in turn, (99%, Aldrich). The latter compound is also known as phenyl
dictates the physical and mechanical properties of the glycidyl ether, or PGE, and will be referred to in the text by its
cured product. For that reason, a fundamental under- acronym.
standing of cure is the first prerequisite for the studies of B. Sample Preparation. Stoichiometric amounts of aniline
modeling and the optimization and design of processing and PGE were mixed at room temperature and stored in dry ice
of composites. for a maximum of 72 h. The reaction kinetics were studied in
Many a researcher studied the mechanism and kinetics the temperature range 90-120 °C. Forty milliliters of the sample
of epoxy-amine reactions, and several fine reviews of the was placed in a 100-mL three-neck bottle, which was immersed
in an oil bath and purged with purified nitrogen. A reflux
subject have been published,2-5 including a recent one with condenser was attached to prevent the evaporation of aniline.
630(!) references.6 But despite the wealth of information
in the literature, the subject of epoxy-amine systems has Temperature was controlled with an Omega CN2010 controller.
At desired time intervals, small aliquots of the reaction mixture
continued to attract considerable scientific efforts, pri- were removed and quenched, prior to the subsequent analysis.
marily because of the contradictory findings and unre- C. Techniques. 1. High-Performance Liquid Chroma-
solved questions regarding some crucial aspects of the
tography (HPLC). High-performance liquid chromatography
epoxy-amine reactions. The principle uncertainties arise (HPLC) analysis was conducted with a Perkin-Elmer LCI-
from the conflicting results concerning the following laboratory computing integrator equipped with a reverse-phase
questions: (1) Is the rate of secondary amine-epoxy C18 column. Gradient elution of the methanol-water system
reaction affected by the preceding reaction between was utilized at a constant flow rate of 1.75 mL/min for 20 min.
primary amine and epoxy (i.e., is there a substitution The solvent had been purged with helium, and a Perkin-Elmer
effect)? (2) Is the reactivity ratio (ratio of secondary LC 235 ultraviolet detector set at 254 nm was used to monitor
the concentration of the species.
amine-epoxy to primary amine-epoxy reaction rate con-
stants) a function of temperature and/or degree of cure? Typically, 5-15 mg of a sample was combined with 10-15 mL
(3) What is the ratio of noncatalyzed and autocatalyzed of methanol. About 20 #tL of the sample was injected. The
rates (or rate constants) for various reaction paths in the concentrations of aniline, PGE, secondary amine (SA), and
overall mechanism of epoxy-amine cure? (4) What are tertiary amine (TA) were calculated from peak areas using
calibration constants established for the pure compounds. TA
the conditions under which etherification, homopolym- was prepared by reacting stoichiometric amounts of aniline and
erization, and intramolecular cyclization become
*
important PGE for 15 h at 120 °C. SA was made by reacting an excess of
in epoxy-amine cure? aniline with PGE for 12 hat 80 °C, after which time excess aniline
At Polytechnic, we have recently embarked upon a was distilled off. Progress of reaction was monitored from chro-
comprehensive research program aimed at answering these matograms taken at different times, as shown in Figure 1.
2. Titration. The argentimetric method was used to follow
* To whom correspondence should be addressed. the concentration of epoxy groups. This method involves hy-
7
Currently with Exxon Corp. drochlorination reaction followed by a measurement of the
1 In the final year of Ph.D. Dissertation work. consumption of chloride ions.

0024-9297/92/2225-0979$03.00/0 © 1992 American Chemical Society

980 Mijovic et al. Macromolecules, Vol. 25, No. 2, 1992

terms of the rate of disappearance of epoxy groups:

-d(E)/dt =
fex(E)(PA) + fe1'(E)(PA)(XH) +
fe2(E)(SA) + V(E)(SA)(XH) + fe3(E)(OH) +
V(E)(OH)(XH) + fc4(E)(E) (1)
The terms in parentheses in eq 1 are concentrations of
different components, and k's are the reaction rate
constants. Subscripts 1,2,3, and 4 refer to epoxy-primary
amine, epoxy-secondary amine, epoxy-hydroxyl, and
epoxy-epoxy reactions, respectively. The prime denotes
catalyzed reactions. Since eq 1 is completely general,
specific modifications must be made whenever the situ-
ation requires. For example, a more complicated scheme
ensues when intramolecular cyclization becomes effective,
which is important in tetrafunctional epoxy formulations
of the type used in advanced composites. A fine treatment
of this subject appears in a recent series of papers by two
groups in Czechoslovakia22 23 and by Pyun and Sung.24
Fortunately, equation 1 can also be simplified by a judicious
choice of components and reaction conditions. First, epoxy
homopolymerization is reported to occur only in the
presence of Lewis base (or acid) type catalysts25·26 and can
be neglected in their absence. Second, the importance of
the etherification reaction between epoxy and hydroxyl
groups has not been firmly established, and, more often
than not, it is said to occur only at high temperatures and
degrees of cure27·28 or in the presence of a large excess of
epoxy groups.29·30 When those conditions prevail in epoxy-
amine mixtures, the rate of disappearance of epoxy groups
can be rewritten in the form analogoues to that proposed
by Horie and co-workers:31
-d(E)/dt =
fet(E)(PA) + fc/(E)(PA)(XH) +
fe2(E)(SA) + fc2'(E)(SA)(XH) (2)
Several authors4·29·32·33 assumed that the epoxy-amine
reactions could be described exclusively by the autocat-
alytic path, in which case the following equation is
obtained:
Figure 1. HPLC chromatograms for the PGE/aniline reaction
at 100 °C for (a) 2, (b) 30, and (c) 80 min. -d(E)/dt =
A1,(E)(PA)(XH) + fe2'(E)(SA)(XH) (3)
Dusek et al.29 reported good agreement between experi-
III. Results and Discussion mental and theoretical results using the following rate
expression:
A. Model Development. Since the early work on the
-d(E)/dt =
fe2(E) (SA)2 + k2'(E) (SA) (OH) (4)
mechanism of epoxy-amine reactions by Shechter et al.,18·19
Chapman et al.,20 and Smith,21 many a researcher has The second-order dependence of the epoxy consumption
contributed to this field of study, and several excellent rate on the concentration of secondary amine suggests
reviews, cited in the Introduction, have appeared in the
that in the absence of a proton donor the amine has a dual
role as a participant in the formation of an intermediate
literature. We shall make no attempt to write another
complex with the epoxy and as a nucleophilic reagent which
review; instead, we shall briefly recap the current under- reacts with the epoxy-amine complex.
standing of the mechanism of epoxy-amine reactions and
then focus our attention to the modeling scheme put We begin our analysis by parlaying the above findings
forward in this work. It is generally agreed that, under into the development of our model, as described below.
The use of model compounds allowed us to simplify the
appropriate kinetic conditions, the following reactions are analysis by introducing two important assumptions: (1)
possible in an epoxy-amine mixture: primary amines due to the chemical structures of the model compounds
(PA)-epoxy (E), secondary amine (SA)-epoxy, hydroxyl (i.e., monofunctionality), intramolecular cyclization is
(OH)-epoxy, and epoxy-epoxy. Furthermore, it is also inconsequential, and (2) in the absence of Lewis base (or
well-known that compounds containing oxygen-hydrogen acid) type accelerators and in the temperature range
or nitrogen-hydrogen bonds (abbreviated here as XH) employed in this study, etherification (epoxy-hydroxyl)
can act as catalysts for epoxy-amine and epoxy-hydroxyl and homopolymerization (epoxy-epoxy) reactions can be
reactions. Since every one of those reactions yields a new neglected. Both assumptions were borne out by the
hydroxyl group, the autocatalytic path in epoxy-amine experimental evidence acquired from HPLC and FTIR
mixtures must be taken into account. On the basis of the studies.
above mechanism one can proceed to write the following Conceptually, our analysis is based upon the existence
general equation for the rate of epoxy-amine reactions in of a transition complex, whose formation is the rate-
Macromolecules, Vol. 25, No. 2, 1992 Epoxy-Amine Kinetics. 1 981

a)
epoxy-amine complex H ...· -
N -

'

O' I

(E . . . PA) 0 K / \ H
or / \ E + PA im----- c -
c -

(E . . . SA) c -
c -

I i

i
-
N -
H -

epoxy-hydroxyl complex H -
0
0 O'
/ X ka
(E . . . OH) 0
/ \ C -
C- + PA -
c -

ch2
-
c -
c -

I I

-
N -

amine-hydroxyl complex H -
0 H

(PA ...
OH) -
N -
H
or l
-
N -

(SA . . .
OH) H .

OH
Figure 2. Three types of hydrogen bond complexes in epoxy- fast I

amine systems. C -

CH2 + amine
i

controlling step in the reaction between an epoxy and an

-
N -
H -
N-
I

amine. This mechanism requires an initial interaction H

between epoxy or amino groups with proton donor

b)
molecules, leading to the formation of a hydrogen bond
complex. Proton donor molecules may be hydroxyl- or H 0
- -
,

amine-containing moieties or impurities present in the K /

O'
\
reaction mixture. Upon a careful examination of epoxy- E + OH c -
c-
amine reaction mechanisms, we have identified the
following three types of hydrogen bond complexes which
can form and are likely to participate in the reactions: (1) .0 -
H H 0

epoxy-amine, (2) epoxy-hydroxyl, and (3) amine-hydrox- /

0
\ '
0
K '
yl. They are schematically presented in Figure 2. Other -
C -
C -
PA --
C -

CH2

complexes, such as epoxy-epoxy and amine-amine may

form but can be neglected in the kinetic analysis, as -
N
I
-
H

successfully argued by several researchers.3,33 Once formed, H

a hydrogen bond complex will next react with an amine, -

0
the latter being available either as a separate amine 0 OH
molecule or as an amine-hydroxyl complex. Reaction / \ I
fast
between a hydrogen bond complex and PA or SA will lead C CH,“
-

hydroxyl
-

r
-
C -

CH2 +
I ;

to SA or TA, respectively, in both cases via an intermediary -

N H N
transition complex, as indicated by eq 5: I
H
I

(hydrogen bond complex) + (amine) -* C)

(transition complex) -*· SA or TA (5) PA + OH
K
—
PA ... OH

Within the general path described by eq 5 we have

identified three specific routes leading to the formation E + OH
K
—
E ... OH
of secondary or tertiary amines. These include route a,
the reaction between an epoxy-amine complex and an
amino group, route b, the reaction between an epoxy- -
0 -
-
0 -

hydroxyl complex and an amino group, and route c, the 0

/ \
kc
/
O'
\
reaction between an epoxy-hydroxyl complex and an -
C c -
N -
H -
-
c -

ch2
amine-hydroxyl complex. A schematic presentation of t
these three routes is shown in Figure 3, and a brief -
0 -
-
N -
H

discussion of their salient features follows. -

O

The first route (route a, Figure 3a), between an epoxy- -

0 -

amine complex and an amino group, was investigated by

0 OH
Arutyunyan et al. ,34 who suggested that amino groups may / \ fast I
act both as electrophilic groups in forming epoxy-amine c c hydroxyl
- -

ch2 -
+
ch2
complex and as nucleophilic groups that react with the
complex to yield secondary or tertiary amines. Arutyu- i
N H
- -

nyan et al. worked with highly purified systems, and their

- -

mechanism is important as it accounts for epoxy-amine 0

- -

reactions in the absence of hydroxyl groups. Nonetheless, Figure 3. Three types of intermediary transition complexes in
this mechanism is significantly only in the earliest stages epoxy-amine systems.
of reaction, where the initial concentration of hydroxyl =
groups remains low. Equations 6 and 7 describe this mech- d(TA)a/dt Wj)e2a(SA)(E-SA) =

anism in terms of the rate of consumption of primary amine W ^28·^ ···8 (8 ) 2(E) (7)
and the rate of production of tertiary amine, respectively: where is the kinetic rate constant for route a, K is the
k&
equilibrium constant, and E—PA and E—SA are epoxy-
-d(PA)a/dt =
Wtfela(PA)(E-PA) = primary amine and epoxy-secondary amine hydrogen bond
complexes, respectively. Subscripts 1 and 2 refer to the
W1AlaKe...PA(PA)2(E) (6) primary amine reactions (eq 6) and tertiary amine reactions
982 Mijovic et al. Macromolecules, Vol. 25, No. 2, 1992

(eq 7), respectively. The weight factor W\ is 0 when the mmole Igr [PGE.Anlllne) rrmole/gi (SAJA]
consumption of epoxy groups by this mechanism is
negligible and 1 when it is not.
The second reaction route (route b, Figure 3b) involves 1.5

interactions between an epoxy-hydroxyl complex and an

amino group and is described by the following general 1

equations:

W2J^khli( PA)(E-XH)
= = 0.5
-d(PA)b/d t

^2^^ ·^ ··· ( A)(XH)(E) (8)

0
= = 0 40 80 I20 160 200
d(TA)b/dí W2]Tfcb2i(SA)(E...XH) TIME (min|
W2£Vb2iE...XH(SA)(XH)(E) (9) •
PGE + Aniline * SA ° TA
where summation is made over i electrophilic groups and
Figure 4. Concentration of components as a function of reaction
other symbols are as defined earlier. The weight factor time in the PGE/aniline system at 100 °C.
W2 is equal to 1 when this route is engaged and 0 when
it is not. For controlled systems, such as ours, where the secondary or tertiary amines. It is important to emphasize
only electrophilic groups are the traces of initially present here that this reaction path has not been hitherto reported
hydroxyl groups and the hydroxyl groups formed in the in the studies of epoxy-amine kinetics. Upon inclusion
course of the reaction, eqs 8 and 9 reduce to a more of route c in the model we were able to arrive at the excellent
convenient form: agreement between predicted and experimental results
observed in this study.
-d(PA)b/dí =
W2(PA)(E)(fcbl + V(OH)) (10) Assuming that the amino groups are depleted (and/or
formed) via all three routes (a, b, and c), one can write the
d(TAydt W2(SA)(E)(feb2 + feb/(OH)) (11)
=
following equations for the rate of disappearance of
Rate constants with primes describe reactions involving primary amine or the rate of production of tertiary amine:
hydroxyl groups produced during reaction, while rate
constants without primes describe reactions involving the
initially present hydroxyls. The concentration of the -d(PA)/df =
-[d(PA)a/dt + d(PA)h/dt + d(PA)c/dt]
^
initially present hydroxyl groups is comprised in the value
of the reaction rate constant. We note that eqs 10 and 11 d(TA)/df d(TA)a/df + d(TA)b/df + d(TA)c/df
=
(18)
are de facto analogous to eq 2 derived by Horie and co-
workers. Using eqs 6-16 and substituting for each corresponding
The third reaction route (route c, Figure 3c) is based on term in eqs 17 and 18, one obtains
the interaction between epoxy-hydroxyl and amine-hy-
-d(PA)/d t
=
W1Aal(PA)2(E) + W2(PA)(E)[febl +
droxyl complexes. The following equations describe the
reaction sequence: fcbl'(OH)] + W3Ac1(PA)(E)(OH)
-

[(PA)/(OH)]/(d(OH)/df) (19)
(PA) + (OH) —
(PA-OH); (PA-OH) =

d(TA)/dt =
W1fea2(SA)2(E) + W2(SA)(E)[fcb2 +
Keql(PA)(OH) (12)
fcb2'(OH)] + WA2(SA)(E)(OH)
-

Keq2
(E) + (OH) *-*
(E—OH); (E—OH) = [(SA)/(OH)]/(d(OH)/di) (20)
Jfeq2(E)(OH) (13) where Aai and fca2 are equal to & * ..· and *2*Ke~sa in
*1 eqs 6 and 7, respectively. Equations 19 and 20 represent
(PA—OH) + (E—OH) -* (SA) + hydroxyl (14) the mechanistic model for epoxy-amine cure kinetics used
in this study.
From eqs 12-14 one can proceed to write an expression for B. Analysis of Results. 1. Reaction Rates. A
the rate of disappearance of amine-hydroxyl complex as typical result of an HPLC experiment is shown in Figure
follows: 4, where concentrations of aniline, PGE, secondary amine
d(PA—OH)/df = ^(OH) d(PA)/df + (SA), and tertiary amine (TA) are plotted as a function
of reaction time at 100 °C. Analogous results were
PA) d(OH)/dt = ^(PA—OH) (E—OH) (15) generated at other temperatures, and repeated runs were
Equations 12-15 can now be combined to obtain the conducted to establish a high degree of reliability and
following expression for the rate of disappearance of reproducibility of data. Mass balances were calculated
primary amine by route c: continually throughout the reaction course, and the
observed deviation never exceeded 5 %. The accuracy of
d(PA)/dt = A1Aeq2(PA)(E)(OH) -

the results was also verified by comparing HPLC and ti-

[(PA)/(OH)] d(OH)/df = fecl(PA)(E)(OH) - tration data, as shown in Figure 5. The computational
analysis of data was conducted with the simusolv modeling
[(PA)/(OH)] d(OH)/df (16) and simulation software (available from Dow Chemical
The rate of production of tertiary amine by route c can Co. and MGA Inc., Concord, MA), which is designed for
be derived by the analogous approach. The main feature development of mathematical models of dynamic systems.
of this route is that it goes into effect only after the hy- To discriminate between possible models, the program
droxyl concentration, which increases during reaction, has uses the maximum likelihood function; the larger it is, the
exceeded a certain critical value. Beyond that point, amine better the fit. The mathematical model used in this work
and hydroxyl groups interact to form a complex, which was written in Advanced Continuous Simulation Language
then reacts with epoxy-hydroxyl complex to produce (ACSL) and is described in detail elsewhere.33
Macromolecules, Vol. 25, No. 2,1992 Epoxy-Amine Kinetics. 1 983

DEGREE OF CURE mmole/ml

HPLC |I2DC) + TITRATION |I20C| * HPLC [HOC]

*
Aniline (exp.) + Aniline (comp.) * PGE (exp.) ° PGE (comp.)

a «
x SA (exp.) o SA (comp.) A TA (exp.) TA (comp.)
TITRATION (HOC] HPLC |IOOC) 0 TITRATION (I00C|

Figure5. Comparison of HPLC and titration results for the Figure 7. Comparison between model prediction and experi-
PGE/aniline system at various temperatures. mental results for PGE/aniline reactions at 110 °C.
mmolelm)

Aniline (exp.) + Aniline (comp.) · PGE (exp.) n PGE (comp.)

x SA (exp,) t SA (comp,) TA (exp.) * TA (comp,)
•
90C + 100C * HOC O I20C

6. Reduced rate of production of tertiary amine

Figure 8. Comparison between model prediction and experi-
Figure as a
mental results for PGE/aniline reactions at 120 °C.
function of hydroxyl concentration.
Table I
As stated earlier, our results have clearly shown that Kinetic Parameters of the Model
the reaction path termed route c must be incorporated in T, eC
the model, although several previous investigators have
L2/(mol2 min) 90 100 110 120 £„ cal/mol
suggested that the epoxy-amine reaction mechanism could
,

be adequately described solely by route b, i.e., the reaction Aib' 0.00304 0.0055 0.0073 0.0085 9.5
A2b' 0.0019 0.0023 0.0026 0.0027 2.8
between an epoxy-hydroxyl complex and an amino group,
Aab7 Aib' 0.31 0.22 0.21 0.16
represented by eqs 10 and 11. The validity of that Aic' 0.0065 0.0087 0.0123 8.8
assumption can be checked by rewriting eq 11 as A2c' 0.0002 0.0050 0.0074 0.0110 12.7
^2c7 klc 0.38 0.42 0.45
d(TA)b/dt(SA)(E) kb2 + Ab/(OH) =
(21)
and plotting the reduced rate of production of tertiary activation energies, obtained from Arrhenius-type plots,
amine (left-hand side of eq 21) as a function of hydroxyl are also listed. Another entry in this table is the reactivity
concentration, as shown in Figure 6. It is clear from this ratio, defined as kn'/kn', an important kinetic parameter
figure that the assumption is correct only below the critical which influences the morphology and properties of net-
value of hydroxyl concentration, which, at the conditions works. The subscript i implies that the reaction can
of this study, is about 2 mmol/mL. Above that value, the proceed by either route b or route c, each of which is
rising importance of amine-hydroxyl complexes necessi- characterized by its own reactivity ratio. In essence, the
tates that route c be included in the model. At the reactivity ratio is a measure of the effect of consumption
conditions of this study, the critical value of hydroxyl of primary amine hydrogen atom on the reactivity of
concentration was found to be independent of tempera- secondary amine hydrogen atom on the same amino group.
ture. A comparison between model predictions and This effect is also known as the substitution effect. A
experimental results is given in Figures 7 and 8. An negative substitution effect signifies a decrease in the
excellent agreement is apparent. reactivity of the secondary amine hydrogen atom, while
2. Reaction Rate Constants. The following approach its absence (i.e., no substitution effect) implies an equal
was applied in the calculation of the kinetic rate constants reactivity of primary and secondary amine hydrogens. The
associated with various reaction routes in the overall kinetic importance of the substitution effect is perhaps best
scheme. Route a was assumed negligible and the corre- conveyed by considering its implications regarding the
sponding rate constants were omitted from the analysis. network morphology. In the absence of substitution effect,
The statistical error in calculating the rate constants for the equal reactivity of all amine hydrogens would favor
route b reactions involving the hydroxyl groups initially the formation of a randomly cross-linked network. How-
present in traces of impurities (i.e., Aib and A2b) was deemed ever, when primary amine hydrogens are much more
too high, and hence their values were considered unreliable. reactive than their secondary amine counterparts, one
The reproducibility and accuracy of the remaining con- would intuitively expect an initial predominance of linear
stants shown in Table I, which include k\b', kib, k\c' and chains followed by subsequent cross-linking, resulting in
kic, were remarkable. The corresponding values of a different network morphology and hence properties. The
984 Mijovic et al. Macromolecules, Vol. 25, No. 2, 1992
Table II
Reported Studies of the Relative Reactivity Ratio
system reacn medium/temp (K) kz/ki method of anal. ref
TGEB/DDS bulk/413-453 0.5 UV-vis 46
DGEBA/DDS bulk/413-453 0.5 UV-vis 34,46
DGEBA/jV-methylaniline bulk/373,425 0.5 GPC 41
BADGE/DDS bulk/383-473 0.4 DSC 28
TGMEDA/DGA 4,4'-DDS, 3,3'-DDS bulk/413,433 0.5 DSC 42
TGE/p-methoxyaniline ethanol/333 0.4 or 0.29 titration 37
TGE/p-nitroaniline ethanol/333 0.62 or 0.26 titration 37
TGE/p-toluidine ethanol/333 0.24 titration 37
TGE/aniline ethanol/333 0.25 titration 37
TGE/p-cyanoaniline ethanol/333 0.24 titration 37
TGE/p-chloroaniline ethanol/333 0.24 titration 37
TGE/p-bromoaniline ethanol/333 0.25 titration 37
TGE/p-iodoaniline ethanol/333 (0.1-0.4 mol/L) 0.30 titration 37
TGE/benzidine ethanol/333 (0.1-0.4 mol/L) 0.31 titration 38
TGE/phenylenediamine ethanol/333 (0.1-0.4 mol/L) 0.35 titration 38
TGE/4,4'-diaminodiphenylmethane ethanol/333 (0.1-0.4 mol/L) 0.33 titraton 38
TGE/4,4'-diaminodiphenyl sulfone ethanol/333 (0.1-0.4 mol/L) 0.52 titration 38
DGEBA/TMAB bulk/373-433 0.16-0.33 FTIR 49
PGE/MPDA THF/343 1 ±0.5 GPC 32
RDGE/aniline THF/343 1 ± 0.5 GPC 32
DGEBA/Tonox bulk/297-338 0.1-0.2 FTIR 45
DGEBA/DDS bulk/450 0.1 FTIR 44
4-methylphenylglycidyl ether/benzy lamine bulk/373 0.42 NMR 47
PGE/aniline o-dichlorobenzene/363-403 0.25 titration 34
BADGE/DDM bulk/353 0.2 titration FTIR 39
TGDDM/DDS bulk/423-450 0.05-0.06 DSC, FTIR 48
DGER/2,6-diaminopyridine bulk/360 1.12 43
DGEBA/MDA excess amine, bulk/333-423 0.08-0.14 titration 36

value of the reactivity ratio, and hence the presence or of equations we were able to derive all significant kinetic
absence of the substitution effect, constitutes a major parameters. The observed agreement between the model
unresolved issue in studies of epoxy-amine kinetics. At predictions and the experimental results was superb. It
this point, it is instructive to peruse Table II, where the was unambiguously shown that the kinetic mechanism is
reported values of reactivity ratios for various epoxy- characterized by a temperature-dependent negative sub-
aromatic amine systems are summarized and seen to stitution effect.
vacillate by more than an order of magnitude. We note
the absence of subscript i on the rate constants in Table References and Notes
II, because the hitherto reported studies of epoxy-amine (1) Gastan, P. Swiss Patent 211,116, 1940.
reaction mechanisms have not accounted for reaction
(2) Barton, J. Adv. Polym. Sci. 1985, 72, 111.
routes other than the one corresponding to route b in this (3) Rozenberg, B. A. Adv. Polym. Sci. 1985, 75,113.
study. At this point, it is instructive to stress that the (4) Oleinik, E. F. Adv. Polym. Sci. 1985, 80, 50.
value of the reactivity ratio can be expressed in terms of (5) Dusek, K. Adv. Polym. Sci. 1985, 78,1.
the reactivity of either amino groups or hydrogen atoms (6) Tanaka, Y.; Bauer, R. S. In Epoxy Resins Chemistry and
on the amino group. In this study, a value of kn'/kw' = Technology·, May, C. A., Ed.; Marcel Dekker: New York, 1988.
0.5 signifies the absence of substitution effect. The relative (7) Kamal, M. R.; Sourour, S.; Ryan, . E. Soc. Plast. Eng. Tech.
Pap. 1973,19, 187.
reactivity values in Table II were normalized accordingly. (8) Kamal, M. R. Polym. Eng. Sci. 1974,14, 231.
Our results of the reactivity ratio for route b (i.e., feab'/kib') (9) Ryan, . E.; Dutta, A. Polymer 1979, 20, 203.
clearly reveal a negative substitution effect, which is in (10) Dutta, A.; Ryan, . E. J. Appl. Polym. Sci. 1979, 24, 635.
agreement with some authors (e.g., refs 34, 36-39,43,44, (11) Barton, J. M. Br. Polym. J. 1979, 11, 115.
and 47-49) and disagreement with others (e.g., refs 24,28, (12) Barton, J. M. Polymer 1980, 21, 603.
41,42, and 46). In addition, there is an obvious temper- (13) Flammersheim, H. J.; Horhold, . H.; Bilstedt, K.; Klee, J.
Makromol. Chem. 1983,184, 113.
ature effect on the reactivity ratio, a fact that must be
(14) Mijovic, J.; Kim, J.; Slaby, J. J. Appl. Polym. Sci. 1984, 29,
considered in modeling studies but, unfortunately, has 1449.
been neglected until recently.46 In the later stages of (15) Mijovic, J. J. Appl. Polym. Sci. 1986, 31,1177.
reaction, when route c comes into play, a weak negative (16) Moroni, A.; Mijovic, J.; Pearce, E. M.; Foun, C. C. J. Appl.
substitution effect is observed, as the ^c'/^ic' ratio Polym. Sci. 1986, 32, 3761.
decreases with decreasing temperature. (17) Mijovic, J.; Wang, . T. J. Appl. Polym. Sci. 1988, 37, 2661.
(18) Shechter, L.; Wynstra, J. Ind. Eng. Chem. 1956, 48, 86.
(19) Shechter, L.; Wynstra, J.; Kurkjy, R. P. Ind. Eng. Chem. 1956,
IV. Conclusions 48, 94.
We have successfully completed a fundamental inves- (20) Chapman, N. B.; Isaacs, N. S.; Parker, R. E. J. Chem. Soc. 1959,
1925.
tigation of epoxy-amine reactions using model compounds. (21) Smith, I. T. Polymer 1961, 2, 95.
A stoichiometric mixture of aniline and l,2-epoxy-3-phe- (22) Matejka, L.; Dusek, K. Macromolecules 1989, 22, 2902, 2911.
noxypropane was prepared and studied in the tempera- (23) Podzimek, S.; Dobas, I.; Svestka, S.; Eichler, J.; Rubin, M. J.
ture range 90-120 °C. Experimental results were obtained Appl. Polym. Sci. 1991, 42, 791, 795.
from high-performance liquid chromatography. A mech- (24) Pyun, E.; Sung, C. S. P. Macromolecules 1991, 24, 855.
anistic reaction scheme was put forward, involving the (25) May, C. A., et al. In Chemorheology of Thermosetting Polymers·,
initial formation of hydrogen bond complexes which, via May, C.A., Ed.; ACS Symposium Series 227; American Chemical
Society: Washington, DC, 1983.
the intermediary transition complexes, lead to the reaction (26) Byrne, C. A.; Hagnauer, G. L.; Schneider, N. S. Polymer 1983,
products by three major routes. From the proposed set 4, 206.
Macromolecules, Vol. 25, No. 2, 1992 Epoxy-Amine Kinetics. 1 985

(27) Gupta, A.; Cizmecioglu, M.; Coulter, D.; Liang, R. H.; Yavrouian, (38) Eichler, J.; Dobas, I. Collect. Czech. Chem. Commun. 1973,38,
Y.; Tsay, F. D.; Moacanin, J. J. Appl. Polym. Sci. 1983, 28, 3461.
1011. (39) Arutyunyan, K. A.; Tonoyan, A. O.; Davtyan, S. P.; Rozenberg,
(28) Riccardi, C. C.; Williams, R. J. J. J. Appl. Polym. Sci. 1986,32, B. A.; Enikolopyan, N. S. Dokl. Akad. Nauk SSSR 1974, 214,
3445. 832.
(29) Dusek, K.; Bleha, M.; Lunak, S. J. Polym. Sci., Polym. Chem. (40) Lunak, S.; Dusek, K. J. Polym. Sci., Polym. Symp. 1975, 53.
Ed 1977 15 2393. (41) Charlseworth, J. J. Polym. Sci., Polym. Chem. Ed. 1980, 18,
(30) Bokare, U. M.; Gandhi, K. S. J. Polym. Sci., Polym. Chem. Ed. 621.
1980, 18, 857. (42) Zukas, W. X.; Schneider, N. S.; MacKnight, W. J. Polym. Mater.
(31) Horie, K.; Hiura, H.; Sawada, M.; Kambe, H. J. Polym. Sci., Sci. Eng. 1983, 49, 588.
Polym. Chem. Ed. 1970, 8, 1357. (43) Rozenberg, B. A. Fiber Sci. Technol. 1983,19, 77.
(32) Kogarko, N. S.; Topolkarayev, V. A.; Trofinova, G. M.; Ivanov, (44) Morgan, R. J. Ado. Polym. Sci. 1985, 72, 1.
V. V.; Berlin, A. A.; Novikov, D. D.; Enikolopyan, N. S. Vysoko- (45) Golub, M. A.; Lerner, N. R.; Hsu, M. S. J. Appl. Polym. Sci.
mol. Soedin. 1978, A20 (4), 756. 1986, 32, 5215.
(33) Vladimirov, L. V.; Artemenko, S. A.; Ivanov, V. V.; Zelenetkii, (46) Sung, C. S. P.; Pyun, E.; Sun, H.-L. Macromolecules 1986,19,
A. N.; Oleinik, E. F.; Salamantna, O. B. Vysokomol. Soedin. 2922.
1980, A22 (1), 225. (47) Borbely, J.; Helen, T. In Crosslinked Epoxies; Sedlacek, B.,
(34) Arutyunyan, Kh. A.; Tonoyan, A. O.; Davtyan, S. P.; Rozen- Kahovec, J., Eds.; Walter de Gruyter: Berlin, 1987.
berg, B. A.; Enikolopyan, N. S. Vysokomol. Soediu. 1975, A17 (48) Chiao, L. Macromolecules 1990, 23,1286.
(8), 1647. (49) Wang, X.; Gillham, J. K. Polym. Mater. Sci. Eng. 1990,63,505.
(35) Fishbain, A.; M.S. Thesis, Polytechnic University, 1991.
(36) Bell, J. P. J. Polym. Sci. 1970, A-2, 417.
(37) Dobas, I.; Eichler, J. Collect. Czech. Chem. Commun. 1973,38, Registry No. Aniline, 62-53-3; l,2-epoxy-3-phenoxypropane,
2602. 122-60-1.