вход по аккаунту



код для вставкиСкачать
Polymer International
Polym Int 49:377±381 (2000)
Glass transition temperature of methyl
methacrylate–ethyl a-benzoyloxymethylacrylate
M Fernández-Garcı́a, R Cuervo-Rodriguez and EL Madruga*
Instituto de Ciencia y Tecnologı́a de Polı́meros (CSIC), Juan de la Cierva 3, 28006-Madrid, Spain
Abstract: Poly(ethyl a-benzoyloxymethylacrylate) (EBMA) and copolymers of methyl methacrylate
(MMA) with EBMA have been prepared by free radical polymerization. Monomer precursors of ethyl
a-benzoyloxymethylacrylate have likewise been polymerized. Glass transition temperatures (Tg) of
homo and copolymers have been determined by differential scanning calorimetry. The Johnston
equation, which considers the in¯uence of monomeric unit distribution on the copolymer glass
transition temperature, has been used to explain the Tg behaviour. Tg12 has been calculated by the
application of the Johnston equation, which gave a value markedly lower than the average value
expected from the additive contribution of the Tg of the corresponding homopolymers.
# 2000 Society of Chemical Industry
Keywords: glass transition temperatures; poly(ethyl a-benzoyloxymethylacrylate); poly(methyl methacrylate);
poly(ethyl methacrylate); methyl methacrylate±ethyl a-benzoyloxymethylacrylate copolymers
It is well known that polymer properties are controlled
by molecular properties such as molecular weight,
molecular weight distribution, chemical composition
and stereochemical distribution of copolymer chain,
and degree of crosslinking, which in turn are a
re¯ection of the kinetic history of the reactions that
occurred during its formation.
The glass transition temperature of a polymer,
representing the molecular mobility of the polymer
chains, is an important phenomenon that in¯uences its
material properties and potential applications. Various
structural characteristics (eg chain stiffness and
intermolecular forces) in¯uence the glass transition
temperature. The mobility of polymer chains depends
on the possibility of rotation around the backbone
carbon±carbon bonds. This itself is determined by the
structure of the monomer units.
Taking into account that the monomer disposition
in a copolymer chain is determined by kinetic events it
is indispensable to consider not only intermolecular
microstructure (average and cumulative chemical
composition) but also intramolecular microstructure
(sequence distribution), because these parameters play
an important role in the understanding of the relations
between molecular structure and properties.
In this study, the determination of the glass
transition temperature of different homopolymers is
performed, taking into account the nature of the lateral
group in the chain. Moreover, methyl methacrylate±
ethyl a-benzoyloxymethylacrylate copolymers are ana-
lysed considering their dependency on the inter- and
intramolecular structure of the copolymer chain.
The synthesis and puri®cation of ethyl a-benzoyloxymethylacrylate (EBMA) have been described elsewhere.1 Commercial methyl methacrylate (MMA)
and ethyl methacrylate (EMA) were puri®ed by
conventional procedures.2
2,2'-Azobisisobutyronitrile (AIBN) was puri®ed by
successive crystallizations from methanol. Benzene
(Merck) for analysis was used without further puri®cation.
Copolymers were prepared by free radical polymerization of mixtures of both monomers with different
compositions, in benzene at 50 °C. The concentration
of initiator was 5 10ÿ3 mol lÿ1 in all cases, and the
total monomer concentration was 2 mol lÿ1. Copolymerization experimental details and microstructural
characterization have been reported previously.1,3
Homopolymerization of monomers was performed
under the same experimental conditions as copolymerization.
Glass transition temperature
Glass transition temperatures were measured using a
Perkin Elmer DSC/TA7DX, PC series differential
* Correspondence to: EL Madruga, Instituto de Ciencia y Tecnologı́a de Polı́meros (CSIC), Juan de la Cierva 3, 28006-Madrid, Spain
Contract/grant sponsor: Comisión Interministerial de Ciencia y Technologı́a (CICYT); contract/grant number: MAT97-682
(Received 17 May 1999; revised version received 12 November 1999; accepted 29 November 1999)
# 2000 Society of Chemical Industry. Polym Int 0959±8103/2000/$17.50
M FernaÂndez-GarcõÂa, R Cuervo-Rodriguez, EL Madruga
Table 1. Glass transition temperature of
DPn ˆ
Ref 11.
scanning calorimeter with a water circulating system
for temperatures over ambient, and a Perkin Elmer
DSC-2 Data Station 3700 with an intracooler for
subambient conditions. The temperature scale is
calibrated from the melting point of high purity
chemicals (lauric and stearic acids and indium).
Samples (about 10 mg) weighed to 0.002 mg with
an electronic autobalance (Perkin Elmer AD4) were
scanned at 10 deg minÿ1 under dry nitrogen
(20 cm3 minÿ1).
The actual value for the glass transition temperature
Tg was estimated as the temperature at the midpoint of
the line drawn between the temperature of intersection
of the initial tangent with the tangent drawn through
the point of in¯ection of the trace and the temperature
of intersection of the tangent drawn through the point
of in¯ection with the ®nal tangent. The current value is
the average for several measurements realized for each
composition. The values determined according to this
criterion may apparently be higher than those obtained
following other procedures. In our case, this is also due
in part to the heating rate employed (10 deg minÿ1).
Methyl methacrylate (MMA) was copolymerized with
ethyl a-benzoyloxymethylacrylate (EBMA) in benzene
solutions at 50 °C, using AIBN as an initiator and
different monomer mixtures with a molar fraction of
MMA in the feed fMMA ranging from 0.099 to 0.897.
Applying the Mayo±Lewis terminal model to copolymer compositions obtained by 1H NMR, the reactivity
ratios were calculated1 and found to be rMMA = 1.342
and rEBMA = 0.320, respectively.
Furthermore, the overall copolymerization rate
coef®cients were measured from dilatometry, their
values ranging from 1.37 10ÿ4 l1/2 molÿ1/2 sÿ1 for
fMMA = 0.897 to 3.05 10ÿ4 l1/2 molÿ1/2 sÿ1 for
fMMA = 0.099. The overall copolymerization rate coef®cient K, mentioned above is de®ned as4
…2fkd †1=2
where f and kd are the values of initiator ef®ciency and
the initiator decomposition rate constant, respectively.
The average values of propagation and termination
rate constants, which are functions of the monomer
molar fraction in the feed, are quoted as kp and kt,
The instantaneous number-average degree of polymerization for a copolymer obtained either in the
presence or absence of solvent may, in general, be
formally expressed in a way similar to that for
kt …2fkd †1=2 ‰IŠ
where [I] and [M] = [M1]‡[M2] are the initiator and
the overall monomer concentration in the feed.
From the parameters de®ned above and consideration of the values obtained for the overall copolymerization rate coef®cient, we must conclude that the
molecular weight of the copolymer is large enough not
to in¯uence the Tg values.
Homopolymerizations were performed under the
same conditions as copolymerization. Ethyl methacrylate (EMA) was introduced to elucidate the side-group
effect on molecular mobility.
The glass transition temperature, Tg, de®nes the
principal transition of amorphous polymeric materials
and is associated with the onset of long range
segmental motion of the polymer backbone. Many
factors in¯uence the value of Tg, but one of the most
important is molecular ¯exibility, which is affected by
the type of substituents. However, the variation of
magnitude of this effect depends largely on the nature
of the side chain with respect to the main chain.7±10 To
determine the effect of substituents on molecular
¯exibility one has to determine the glass transition
temperatures of related polymers.
Tgs for PMMA, PEMA, PEBMA have been
measured and their values are shown in Table 1 along
with the value previously reported11 for poly(ethyl aTable 2. Glass transition temperatures of EBMA–MMA copolymers
obtained in benzene solution at
fMMA a
Molar fraction of MMA in the feed.
Molar fraction of MMA in the
copolymer chain.
Polym Int 49:377±381 (2000)
Glass transition temperatures of MMA±EBMA copolymers
Scheme 1. Monomer structures.
hydroxymethacrylate), PEHMA (see Scheme 1 for
monomer formulae).
As expected, the Tgs for poly(n-alkyl methacrylate)s
decrease as the chain length of the ester alkyl group
increases. The temperature increment found between
PMMA and PEMA is 43 K. This is in agreement with
the well established criterion that increasing the length
of a ¯exible alkyl side chain brings a monotonic
decrease in the value of Tg in a series of vinyl polymeric
The substitution of a hydrogen atom in the EMA amethyl group for a benzoyloxy group produces an
increase the Tg of about 56 K. The rigid and bulky
benzoyloxy group does not facilitate molecular rotation, and as a consequence the Tg is raised.
Several authors have proposed correlations between
the chemical structure and Tg.13 Their methods are
usually based on the assumption that the structural
groups in the repeating units provide weight additive
contributions to the Tg.
It is worth noting that the introduction of a
methylene group in the ester group of PMMA acts as
a ¯exible spacer and produces a decrease in Tg of 43 K,
while the substitution of hydrogen atom in the amethyl group for a benzoyloxy group produces an
increase of 56 K. As a consequence, the Tg of PMMA
is only 13 K lower than that of PEBMA.
Similar behaviour is observed when the Tg of
PMMA is compared with that of PEHMA,11 which
is the homopolymer of the EBMA precursor monomer. When one hydrogen atom in the PEMA a-methyl
group is changed for a hydroxyl group, producing
PEHMA, an increase in Tg of about 18 K is observed.
This is not only because the a-hydroxyl group is bigger
than a hydrogen atom, but because of the intra±inter
molecular interactions between hydroxyl groups.
These interactions produced by hydrogen bond hinder
the chain movements, increasing the stiffness. Consequently, the difference between PMMA and PEHMA
is 25 K, corresponding to a 43 K decrease because of
methylene group incorporation and an 18 K increase
due to the hydroxyl group.
From those results it can be seen that the Tg of
PEBMA and PEHMA can be approximated by the
proper choice of structural groups, in other words by
choosing an additive contribution of the latter. The
Polym Int 49:377±381 (2000)
difference in Tgs between PEHMA and PEBMA (the
a-hydroxyl group is changed for a a-benzoyloxy group)
is approximately 39 K. In this case, the introduction of
a relatively long chain does not provide greater
¯exibility because the aromatic ring is bulky.
The physical properties of a copolymer are fundamentally determined by its Tg. Originally, copolymer
Tgs were described by simple additive relations,14,15
based on free volume theories15±17 or thermodynamic
theories,14 which did not take into consideration the
sequence distribution of the monomer units and the
effect of their compatibility on steric and energetic
interactions. The free volume theory developed by Fox
and Flory16 suggests that the glass transition occurs
when the free or unoccupied volume of the material
reaches a constant value and does not decrease further
as the material is cooled below its Tg. A thermodynamic theory, proposed by Gibbs and DiMarzio14 is
based on the change of material con®gurational
entropy as a function of temperature. At equilibrium,
it postulates that the con®gurational entropy Sc equals
zero at the glass transition. However, these linear
relationships often failed to predict accurate glass
transition temperature of copolymers, because they
neglected the effects of the chemical nature and
organization of the monomers on the mobility of a
polymer chain.18 Several models were therefore
proposed18±20 that differentiated between homo
(M1±M1, M2±M2) and heterolinkages (M1±M2 or
M2±M1), recognizing the signi®cant effect of monomer arrangement on Tg, so that both negative and
positive deviations from linearity could be predicted.
The relations proposed by Barton,19 Uematsu and
Honda,21 Hirooka and co-workers,22 Furukawa23 and
Suzuki et al 24 may be considered as extensions of the
Gibbs±DiMarzio14 relation, whereas the approach by
Johnston18 is based on the Fox±Flory equation.16 A
third relation, developed by Couchman and Karasz,20
is based on mixed-system entropy and was also able to
predict composition-dependent Tgs for a variety of
Among all of these, those derived by Johnston, Barton
or Couchman which correlate Tg to the dyad distribution in the instantaneous copolymer molecules, exhibit
better agreement with experimental Tgs.25,26 In this
work we are going to use the Johnston equation, which is
M FernaÂndez-GarcõÂa, R Cuervo-Rodriguez, EL Madruga
Figure 2. Glass transition temperature of MMA–EBMA copolymers as a
function of methyl methacrylate weight molar fraction in the copolymer
Figure 1. Plots of the glass transition temperature of MMA–EBMA
copolymers according to the linearized expression of Johnston.18
based on the free volume concept and the inter±
intramolecular composition of the copolymer. The
description of Johnston's model is as follows.
The Johnston equation18 assumes that M1M1,
M1M2 or M2M1 and M2M2 dyads have their own Tg,
with the overall Tg of a copolymer described by the
following expression:
w1 P11 w2 P22 w1 P12 ‡ w2 P21
in which w1 and w2 are the weight fractions of
monomeric units in the main chain, and P11, P12, P21
and P22 are the probabilities of having various linkages
which can be calculated considering the Mayo±Lewis
terminal model by using the monomer feed composition and the monomer reactivity ratios.27 Tg11 and Tg22
are the glass transitions of the respective homopolymers, and Tg12 is the supposed glass transition for the
alternating sequence M1M2 or M2M1.
To apply the Johnston theory, it is necessary to
determine the glass transition temperature Tg12 of a
strictly alternating copolymer. In this case, Tg12 for
MMA±EBMA copolymers is unknown, but can be
calculated from our own experimental values: Tg of
PMMA, Tg of PEBMA (Tg11 and Tg22, respectively)
and Tgs of a copolymers series of varied compositions
obtained at low conversion. A linearized form of eqn
(1) is used to determine Tg12. As can be observed in
Fig 1, the experimental data produce a very good
straight line with the Tg12 value 396.6 K. It is
important to note that, within experimental accuracy,
Tg12 (MMA±EBMA or EBMA±MMA link) has the
same value as Tg11 (MMA±MMA link).
Considering the Tg12 value found and according to
the Johnston equation, the curve of which shows it Fig
2 is drawn showing the relation between Tg and methyl
methacrylate weight molar fraction wMMA in the
copolymer. As can be observed, when wMMA is lower
than 0.15 Tg decreases markedly, then decreases
slowly until wMMA approaches 0.45, and from there
the variation is practically nil. This can be explained as
the in¯uence of the copolymer microstructure on Tg.
From the monomer feed composition, monomer
reactivity ratio values and using Bernoulli's statistic,
it is easy to calculate the formation probabilities of
M1M1, M2M2 and M1M2 or M2M1 dyads as a function
of monomer molar fraction in the feed. Knowing this,
it is simple to determine the dyad molar fraction as a
function of the weight fraction of the monomer unit.
In Fig 3 are represented the calculated dyad molar
fractions for this system. As can be observed, at wMMA
values lower than 0.15 the EBMA±EBMA (22) dyad
concentration is higher than 50%, which means that
the contribution of Tg22 to the overall copolymer Tg is
predominant. For wMMA between 0.15 and 0.45, the
sum of MMA±MMA (11), MMA±EBMA (12) and
EBMA±MMA (21) monotonically increases with
respect to EBMA±EBMA, and the overall copolymer
Tg decreases moderately. For wMMA values higher than
0.45, the EBMA±EBMA dyad is lower than 10% and,
Figure 3. Dyad molar fractions versus methyl methacrylate weight molar
fraction in MMA–EBMA copolymer chains.
Polym Int 49:377±381 (2000)
Glass transition temperatures of MMA±EBMA copolymers
within experimental error, its contribution to the
overall copolymer Tg vanishes.
In contrast, Hirooka and co-workers,22 observed
that the Tg for dyads (Tg12) calculated from the Tgs of a
series of copolymers of varied composition, did not
always correspond with that of the chemically synthesized alternating copolymer. This deviation depends
on the type of Tg±composition relationship of the
statistical copolymer. The Tg of a pure alternating
copolymer should be higher, lower or similar to the Tg
estimated from the Tg±sequence distribution when the
Tg±composition curve for the statistical copolymer is
convex, concave or linear, respectively.
Tonelli28 used conformational entropy as a characterizing parameter for polymer intramolecular chain
¯exibility. Deviations positive, negative or no deviation
from bulk additive, namely Tg12, estimated from Tg±
sequence distribution behaviour are produced when
the conformational entropy for a given copolymer
chain is, respectively, lower than, higher than or
similar to the weighed sum of entropies calculated
for the constituent homopolymer chains.
Moreover, the Tg of the polymer is related to the
chain ¯exibility and this parameter is, to a large extent,
a re¯ection of the rotational barrier about the bond
linking two monomer units. Depending on the
rotational barrier of the heterolink bond being similar
to, higher or lower than the averaged rotational barrier
of the homolink bond, the copolymer Tg±composition
behaviour will be linear, or show positive or negative
deviations from linearity.29 Hirooka and co-workers22
proposed that the difference between the average
Tg (Tg ˆ …Tg11 ‡ Tg22 †=2) and the supposed Tg12 of an
alternating copolymer may be regarded as a measure of
the heterolink stiffness. In this work, Tg is 403.2 K
whereas a lower value of Tg12 (396.6 K) is obtained
using the Johnston equation. This indicates that this
system has a heterolink stiffness lower than that of the
average homopolymer links and, consequently, a
negative deviation from linearity in the Tg±composition plot observed in Fig 2 is expected. The cause of
this behaviour is not clear, because the chain ¯exibility
depends not only on the rotation barrier but also on
the chain packing, side-chain stiffness, dipole interactions, etc.
Taking into account all these features and the small
differences between the Tg of the homopolymers, good
agreement between experimental and theoretical
values is found, indicating that the Johnston equation
and the terminal model of Mayo and Lewis through
reactivity ratios may be used to describe the dependence between the experimental glass transition
temperature of MMA±EBMA copolymers and their
sequence distribution.
It has been found that the differences between the
glass transition temperature of the a-substituted
acrylate homopolymers studied in this work with
Polym Int 49:377±381 (2000)
respect to the Tg of PMMA could be explained
through the additive contribution of the lateral
The Tg of the homopolymer, inter and intramolecular structure, together with the Johnston equation,
allow the experimental variations of the Tg of MMA±
EBMA copolymers to be described.
This research has been supported by the ComisioÂn
Interministerial de Ciencia y TecnologõÂa (CICYT),
MAT97-682. The authors would like to thank
Professor Dr FernaÂndez-MartõÂn from the Instituto
del Frio (CSIC, Madrid) for his advice and for
allowing us to use the equipment at his Institute.
1 Cuervo-RodrõÂguez R, FernaÂndez-Monreal MC, FernaÂndezGarcõÂa M and Madruga EL, Macromol Rapid Commun
16:503 (1995).
2 Stickler M, Makromol Chem Macromol Symp 10/11:17 (1987).
3 Cuervo-RodrõÂguez R, FernaÂndez-Monreal MC, FernaÂndezGarcõÂa M and Madruga EL, J Polym Sci Polym Chem
35:3483 (1997).
4 Fukuda T, Ma YD, Kubo K and Takada A, Polym J 21:1003
5 Hamielec AE, MacGregor JF and Penlidis A, Makromol Chem
Macromol Symp 10/11:521 (1987).
6 Gao J and Penlidis A, J Macromol Sci Rev Macromol Chem Phys
38:651 (1998).
7 Halden RA, Schell WJ and Shima R, J Macromol Sci B-1:759
8 Hoff EAW, Robinson DW and Willbourn AH, J Polym Sci
18:161 (1955).
9 Rogers SS and Mandelkern L, J Phys Chem 61:985 (1957).
10 Bandrup J and Immergut EH, Polymer Handbook, 2nd edn.
Wiley-Interscience, New York (1975).
11 FernaÂndez-GarcõÂa M, Madruga EL, Cuervo-RodrõÂguez R and
FernaÂndez-Monreal MC, Polymer 35:4474 (1994).
12 Reimschussel HK, J Polym Sci Polym Chem Ed 17:2447 (1979).
13 Van Krevelen DW, Properties of Polymers, 3rd edn, Elsevier, New
York (1990).
14 Gibbs JH and DiMarzio EA, J Polym Sci Part A, 1:1417 (1963).
15 Fox TG, Bull Am Phys Soc 1:123 (1956).
16 Fox TG and Flory PJ, J Appl Phys 21:581 (1950).
17 Gordon M and Taylor JS, J Appl Chem USSR 2:492 (1952).
18 Johnston NW, J Macromol Sci Rev Macromol Chem C 14:215
19 Barton JM, J Polym Sci C 30:573 (1970).
20 Couchman PR and Karasz FR, Macromolecules 11:117 (1978).
21 Uematsu I and Honda K, Rep Prog Polym Phys Jpn 8:111 (1975).
22 (a) Hirooka M, Jabunchi H, Iseky J and Nakai Y, J Polym Sci A1
6:1381 (1968); (b) Hirooka M and Kato T, J Polym Sci Polym
Lett 12:31 (1974).
23 Furukawa J, J Polym Sci Polym Symp 51:105 (1975).
24 Suzuki H, Kimura N and Nishio Y, J Therm Anal 46:1011 (1996)
and references cited therein.
25 Fernandez-Garcia M, Gonzalez-Lopez MMC, Barrales-Rienda
JM, Madruga EL and Arias C, J Polym Sci Polym Phys 32:1191
26 Guillot J and Emilie B, Makromol Chem Rapid Commun 12:117
27 Harwood HJ and Ritchey WM, J Polym Sci B 2:601 (1964).
28 Tonelli A, Macromolecules 10:716 (1977).
29 McEven IJ and Johnston AF, in Alternating Copolymers, Ed by
Cowie JM, Plenum Press, New York (1985).
Без категории
Размер файла
96 Кб
Пожаловаться на содержимое документа