Materials Sciences and Applicatio n, 2011, 2, 1033-1040
doi:10.4236/msa.2011.28140 Published Online August 2011 (
Copyright © 2011 SciRes. MSA
Optical Microscopy as a Tool to Correlate
Morphology and Thermal Properties of Extruded
PET/PC Reactive Blends
Luis Claudio Mendes*, Patricia Soares da Costa Pereira
Instituto de Macromoléculas Professora Eloisa Mano-IMA, Universidade Federal do, Rio de Janeiro-UFRJ, Centro de Tecnologia,
Bloco J, Avenida Horácio, Macedo, Rio de Janeiro, RJ, Brasil.
Email: *
Received May 16th, 2011; revised June 20th, 2011; accepted June 30th, 2011.
Catalysed and uncatalysed reactive extruded poly(ethylene terephthalate)/poly(bisphenol-A carbonate) (PET/PC)
blends phase structure at compositional range containing 0 - 100 wt% of both parent polymers were evaluated. Phase
separation was supported by TG/DTG, DMA and DSC. The changes on Tg and Tm of the parent polymers were associ-
ated to the esterification and tra nsesterificatio n rea ction s inside the pha ses and into th e interfac ial reg ion . Acco rd ing to
optical observations, not yet published in this matter, the blend morphology was dictated, either composition or melt
flow rate and in the who le composition range, a matrix-droplet morp hology was no ticed. PET was only ab le to be crys-
tallized in blends in which it was the matrix. Mostly, the PET/PC blends revealed to be partially miscible systems in
which the level of the transesterification/esterification reactions was driven by the kind of matrix. The latter showed
great influence on the thermal properties.
Keywords: PET/PC, Phase Str ucture, Morphology , OM, DMA, DSC , TG/DTG
1. Introduction
It is well known the qualities and significance of poly
(ethylene terephthalate) (PET) and poly (bisphenol-A
polycarbonate) (PC) as commodities and engineering
plastics. In order to gather their individual characteris-
tics to yield a new material, their blends have been
studied for at least a quarter of century. Jointly, chem-
ical similarity, capacity of reacting to each other in the
molten state and fairly easy processing among others
have been attractive to encourage their blends investi-
gation. When dealing with reactive processing, Utracki
[1] underlined that to ascertain the renewal of the in-
terface, ability to react across the interface, sufficient
reaction rate and as well the stability of both the
formed chemical structures and morphology are vital
conditions for its success. The expected miscibility due
to the occurrence of esterification and transesterifica-
tion reactions during the mechanical blending has been
an object of discussion. The discrepances could be at-
tributed to many factors—molar mass, melt flow rate
(MFR), intrinsic viscosity, end groups content of the
homopolymers; temperature, pressure, screw profile, res-
idence time of processing; absence and presence of
catalyst—for instance. In recent work, Al-Jabareen et
al. [2] investigated reactive extruded PC/PET blends—
only PC rich blends, with/without different transesteri-
fication catalysts and with the addition of Irganox as
heat stabilizer. Even considering that the blends were
processed twice in different extruders and that also two
screw speeds were also applied, it was found that all
blends are formed by a PC matrix and PET dispersed
phase. Hay and co-authors [3] prepared PET/PC
blends—PC content 10 - 50 wt%, without and with
lanthanum acetylacetonate, at utmost extrusion tem-
perature. The blends were considered completely im-
miscible over the composition range studied. The rela-
tionship between miscibility and chemical structures of
PET-PC uncatalysed reactive blend prepared by melt
mixing, for prolonged time, was assessed by Zhang and
collaborators [4]. With increasing reaction time, they
detected the progressive enhance of the miscibility un-
til a single Tg has been reached. Melt mixing PET/PC
blends have been yielded by Marchese et al. [5] taking
into account distinguished PET molar mass, the pres-
Optical Microscopy as a Tool to Correlate Morphology and Thermal Properties of Extruded PET/PC Reactive Blends
ence of different residual catalysts in PET and as well
distinct mixing times. They pointed out that the pres-
ence of single or two Tgs were dependent on the mo-
lecular mass of the PET blocks, considering that the
PET and PC crystallizations were possible due to the
chain rearrangements and block length. An article on
reactive extruded PET/PC blends in which PET waste
from recycled bottles was blended with virgin PC,
without/with catalysts, in a whole range of composition
was published by Carrot and coworkers [6]. Their
rheological and SEM observations showed the occur-
rence of matrixdroplet morphology for all compositions
and a phase inversion point for PET/PC 50/50 wt%
blend. Stannous octoate was tested as a transesterifica-
tion inducer in reactive extruded PET/PC blends in a
wide range of composition. According to the authors, it
was found two distinct glass transitions in all cases
confirming that the polymers are not miscible even in
the presence of catalyst [7]. A study on the efficiency
of different transesterification catalysts on phase be-
havior of PET/PC blend (50/50 wt%), by melt mixing
at prolonged reaction times was performed. A single
Tg was observed when the block copolymer length had
reached fifteen monomeric units [8]. Some articles
have been published by Mendes and his group [9-12].
Latterly, two works on rheological study and effec-
tiveness of catalyst on PET/PC reactive extruded
blends were issued. They evidenced the significance of
the matrix for driving the interchange reactions and the
efficiency of the catalyst as a real tranesterification
promoter [13]. In the present work, we tried to corre-
late the morphology and thermal properties of the
PET/PC reactive extruded blends taking into consid-
eration the importance of the blend matrix based on the
progress of the esterification and transesterification
2. Experimental
2.1. Materials
PET and PC manufacturers as well as their characteris-
tics were displayed in Table 1. Commercial cobalt ace-
tylacetonate produced by J. T. Baker Chemical Co. was
used as catalyst.
2.2. Blending
The reactive blend was processed in a co-rotating
twin-screw extruder (L / D = 36, screw diameter = 22
mm), equipped with vacuum system, at temperature
range of 190 - 255˚C and speed of 150 rpm. The blend
compositional range covered 0, 20, 50 and 100 wt% of
each polymer, with (500 ppm) and without cobalt acety-
lacetonate II, a transesterification catalyst. Previously, a
master of PET/catalyst was extruded for preparing cata-
lysed blend. Prior to blending, the polymers were dried at
120˚C, during 8 hours.
2.3. Thermogravimetry/thermogravimetry
Derivative (TG/DTG)
The thermal behavior of the blends were taken from
TG/DTG curves using a TA thermogravimetric analyser
model Q 500, in the temperature range of 30 - 700˚C, at
10˚C·min–1, under nitrogen atmosphere. The onset,
maximum degradation and end temperatures, respec-
tively, Tonset, Tmax, Tend and residue contents were deter-
2.4. Differential Scanning Calorimeter (DSC)
A TA calorimeter model Q1000 was used to register the
calorimetric events in the heating and cooling regimes.
The first regime of heating was performed from 40 to
300˚C at heating rate of 10˚C·min–1, under nitrogen at-
mosphere, kept for 2 min for eliminating the thermal
hystory. After that, the first regime of cooling was ap-
plied up to 40˚C at maximum rate. The second regime of
heating followed the same temperature range and rate of
the first one. Finally, the second regime of cooling was
conducted until 40˚C at rate of 10˚C·min–1. The glass
transition, crystallization, melting temperatures, respec-
tively (Tg), (Tc), (Tm) and enthalpy of fusion (ΔHm) were
measured. The heating and cooling crystallization tem-
peratures, Tch and Tcc, were determined when it was pos-
sible. The PET’s degree of crystallization (Xc) was cal-
culated from the ratio of PET endothermic peak area
(ΔHm) and the enthalpy of fusion of 100% crystalline
PET (136 J·g–1).
2.5. Dynamic-Mechanical Analysis (DMA)
A dynamic-mechanical analyser from TA equipment
Table 1. Polymers origin and characteristics.
Polymer Density ( Melt flow rate (g.10min-1) Manufacturer
PET 1.39 33.0 Mossi & Ghisolfi Group
PC 1.20 2.5 GE Plastics South America
Copyright © 2011 SciRes. MSA
Optical Microscopy as a Tool to Correlate Morphology and Thermal Properties of Extruded PET/PC Reactive Blends 1035
model DMA-298 allowed to assess the Tan δ using rec-
tangular specimens of 40 × 10 × 0.5 mm in the following
experimental conditions: bending mode, 1Hz, tempera-
ture range 50 - 170˚C and heating rate of 2˚C·min–1.
2.6. Morphological Analysis
The blend morphology was featured through a Zeiss mi-
croscopy model THMS 600 from squeezed film between
two microscope glass slices. The assembly was inserted
into the microscope hot stage and heated from 25 to
280˚C where was kept for 2 minutes in order to eliminate
the thermal history. After that, the cooling was per-
formed until 25˚C. The phase separation and the crystal-
lization process were monitored by taking photographs.
3. Results and Discussion
3.1. Thermogravimetry/thermogravimetry
Derivative (TG/DTG)
TG/DTG curves and Tonset, Tmax, Tend, residue content are
arranged in Figures 1-4 and Table 2, respectively. The
homopolymers degraded at a single step, whereas the
blends into two ones for both kinds of blends. The latter
showed the Tonset values below to the PET homopolymer.
The DTG curves showed two peaks of degradation in
which the first one was correlated with the PET-rich pha-
se while the second one with the PC-rich one. It is worth
observing that the derivative weight peak intensity
changed with the composition for the uncatalysed blends
Figure 1. TG without catalyst.
Figure 2. TG with catalyst.
Copyright © 2011 SciRes. MSA
Optical Microscopy as a Tool to Correlate Morphology and Thermal Properties of Extruded PET/PC Reactive Blends
Figure 3. DTG without catalyst.
Figure 4. DTG with catalyst.
Table 2. TG/DTG parameters for PET, PC and blends.
Degradation temperature (˚C) Tmax
Sample Tonset T
Residue content
(%) PET-rich phase PC-rich phase
PET 375 465 12 436 -
PC 460 525 23 - 508
80/20a 318 510 11 443 510
50/50a 320 520 19 445 494
20/80a 322 518 18 445 501
80/20b 325 519 18 431 483
50/50b 309 517 19 447 485
20/80b 325 515 17 429 486
a- without catalyst; b- with catalyst
but its magnitude is always at higher values for PET-rich
phase when the catalysed systems are considered. This
could be ascribed to the unlike level of the acidoly-
sis/alcoholysis and transesterification reactions in both
blends. The percentages of coal residue are between the
homopolymers. Herein, the TG/DTG analysis showed
Copyright © 2011 SciRes. MSA
Optical Microscopy as a Tool to Correlate Morphology and Thermal Properties of Extruded PET/PC Reactive Blends1037
that both PET/PC blends presented at least two phases.
3.2. Dynamic-Mechanical and Calorimetric
Table 3 shows the Tg values from DMA for all blends.
Most of them were shifted in relation to the parent ho-
mopolymers. Meanwhile the Tg of PE-rich phase dis-
placed to higher temperatures the Tg of PC-rich phase
showed tendency to diminish, a typical behavior of par-
tially miscible system. The role of the catalyst as trans-
esterification promoter did not change the PET and PC
Tg’s values of the 80/20 blend. Only the PET’s Tg was
altered in the blend with equal polymers weight percent.
Nevertheless, a single Tg was detected in the 20/80
blend. Based on the Tg values, it was determined the
amount of the polymer/copolymer inside of each phase
(Table 4). The catalyst had a remarkable role in blends
with 50 and 80 wt% of PC. In all cases, DMA results
pointed out that the PET and PC formed miscible blends
Table 5 shows the Tch, Tm and Xc for all blends. In
both kinds of blends, when detected the Tch was shifted
to higher temperatures while the Tm and Xc decreased.
The presence of PC low molar mass and PET/PC co-
polymer inside the PET-rich phase retarded the PET
crystallization and were also able to influence the PET
crystal size and crystallizablity.
3.3. Optical Microscopy Analysis
All blends showed morphology like matrix-droplet in both
the molten state and after cooling besides an interfacial
region (Figures 5 and 6). The droplets were somewhat
distorted due to the coalescent effect with average dimen-
sions in the range of 200-1200 microns. Blend morphol-
ogy may be influenced by several factors – blend compo-
sition, polymers viscosity ratio, interfacial tension, shear-
ing, processing time, among others [14]. For checking
which polymer was the matrix or the disperse phase in the
blends, the PET crystallization was monitored. In the
80/20 blend, the morphology was driven by compositional
effect. As PET crystallization only occurred outside the
disperse domains, it was deduced that the matrix was PET.
For intermediate composition, polymers viscosity ratio
effect predominated to the formation of the blend mor-
phology. The PET crystallization began firstly outside the
disperse phase and after few minutes feeble crystals ap-
peared in it. It led to the conclusion that PET was the ma-
trix. In the blend with the highest PC content, the compo-
sitional effect was uppermost in blend morphology. Since
PET crystals were not observed neither matrix nor disperse
phase it was deduced that PC was the matrix. Even in the
presence of the cobalt catalyst, the PET/PC blends re-
mained as two phase systems. The observations taken from
Table 3. Tg measurements from DMA.
Tg (oC)
Sample PET phasePC phase Tg FOX(oC)
PET 81 - -
PC - 158 -
80/20a 89 147 95
50/50a 94 146 116
20/80a 78 155 140
80/20b 89 148 95
50/50b 107 145 116
20/80b - 140 140
a- without catalyst; b- with catalyst
Table 4. PET and PC content in each phase.
PET/PC blend PC in PET phase (%) PET in PC phase (%)
80/20a 11 4
50/50a 20 6
20/80a 0 2
80/20b 11 4
50/50b 35 7
20/80b --- 12
a- without catalyst; b- with catalyst
Table 5. DSC parameters for PET, PC and blends.
Sample Tch (oC) Tm (oC) Xc (oC)
PET 137 247 20
PC - - -
80/20a 178 234 18
50/50a 172 233 10
20/80a - - -
80/20b 186 229 10
50/50b - - -
20/80b - - -
a- without catalyst; b- with catalyst
optical microscopy may help to understand the ther mal
In general, the exchange reactions in the blend of poly-
esters may proceed by two different mechanisms – esteri-
fication and transesterification. The first one takes place by
a direct attack of reactive chain outer functional groups
(hydroxyl, ester, carboxyl) on inner groups (ester, carbon-
ate). The second mechanism occurs by reactions between
inner functional groups (ester and carbonate) situated
along the polymer chain. Herein, the parent polymers are
high molecular mass materials and it is expected that the
number of skeletal ester groups will be higher than the end
chain functional ones. Then, it is assumed that transesteri-
fication firstly occurs in PET/PC blend.
As the blend is an immiscible system in the molten state,
the inner-inner mechanism (Scheme 1-transesterification)
operates through the interfacial region. Considering the
PET as matrix, there is a sea of PET molecules surrounded
by islands of PC. The attack of PET ester-inner groups on
Copyright © 2011 SciRes. MSA
Optical Microscopy as a Tool to Correlate Morphology and Thermal Properties of Extruded PET/PC Reactive Blends
Figure 5. Uncatalysed blends optical photographs: 80/20 (a-molten state; b-solid state), 50/50 (c-molten state; d-solid state),
20/80 (e-molten state; f-solid state).
Figure 6. Catalyzed blends optical photographs: 80/20 (a-molten state; b-solid state), 50/50 (c-molten state; d-solid state),
20/80 (e-molten state; f-solid state).
Copyright © 2011 SciRes. MSA
Optical Microscopy as a Tool to Correlate Morphology and Thermal Properties of Extruded PET/PC Reactive Blends 1039
PC carbonate-inner groups led to the formation of aroma-
tic-aromatic ester copolymer in the interfacial region and
the dissolution of the aliphatic-aromatic ester copolymer
into the PET-rich phase. Inside the latter, the PC segment
inserted along copolymer chain is subjected to trans-
esterification reaction and direct attack of PET reactive
chain outer functional groups-hydroxyl and car-
boxyl-(Scheme 1-acidolysis and alcoholysis). These re-
actions provided copolymer with PC low chain segments
and released PC low molecular mass molecules inside
the matrix. These events are dependent on the amount of
PC domains and may be interrupted until the access of
PET chain to the PC domains is avoided by the presence
of the copolymer in the interfacial region. Thus, the
highest Tg value and the amount of PC in PET-rich phase
were noticed to the 50/50 blend and partially miscible
systems were achieved. Our results differ from those
found by Arefazar et al. [15]. They stated that PET/PC
mixing blends prepared without catalyst were immis-
The kind of matrix seems to influence the occurrence
of the transesterification reaction. As can be seen, the
glass transition temperatures of PET and PC phases did
not change in the blend containing high content of PC
and without catalyst. Although there is a sea of PC mo-
lecules in the matrix the energy level necessary for the
attack of PC carbonate groups on the PET ester ones
seems to be high. Only, in the presence of catalyst the
activation energy was enough reduced allowing the oc-
currence of the transesterification reaction and then, the
displacement of Tg was noticed. Additionally, as stated
by Montaudo et al. [16], PC phenol end groups must be
considered unreactive which led to conclude that the al-
coholysis reaction of the PC hydroxyl end groups on PET
ester linkage in the PC-rich phase could be neglected.
4. Conclusions
The relationship between morphology and thermal prop-
erties of PET/PC melt mixing blends was investigated. It
was observed the interference of the kind of matrix on
the transesterification reaction and its consequence on the
thermal properties. The catalyst was actually effective in
some blends and the systems are partially miscible
5. Acknowledgements
The authors thank the Conselho Nacional de Desenvol-
vimento Científico e Tecnológico (CNPq), Fundação
Scheme 1. PET/PC Alcoholysis, acidolysis and transesterification reaction.
Copyright © 2011 SciRes. MSA
Optical Microscopy as a Tool to Correlate Morphology and Thermal Properties of Extruded PET/PC Reactive Blends
Coordenação do Aperfeiçoamento de Pessoal de Nível
Superior (CAPES) and Universidade Federal do Rio de
Janeiro (UFRJ) for supporting this investigation.
[1] L. A. Utracki, “Compatibilization of Polymer Blends,”
Canadian Journal of Chemical Engineering, Vol. 80, No.
6, 2002, pp. 1008-1016. doi:10.1002/cjce.5450800601
[2] A. Al-Jabareen, S. Illescas, M. L. I. Maspoch and O. O.
Santana, “Effects of Composition and Transesterification
Catalysts on the Physico-chemical and Dynamic Proper-
ties of PC/PET Blends Rich in PC,” Journal of the Mate-
rials Science, Vol. 45, No. 24, 2010, pp. 6623-6633.
[3] Y. Kong and J. N. Hay, “Miscibility and Crystallization
Behavior of Poly(Ethyelene Terephthatale)/Polycarbonate
Blends,” Polymer, Vol. 43, No. 6, March 2002, pp. 1805-
1811. doi:10.1016/S0032-3861(01)00772-8
[4] Z, Zhang, Y. Xie and D. Ma, “Relationship between Mis-
cibility and Chemical Strucutures in Reactive Blending of
Poly(Bisphenol A Carbonate) and Poly(Ethylene Tere-
phthalate,” European Polymer Journal, Vol. 37, No. 10,
October 2001, pp. 1961-1966.
[5] P. Marchese, A. Celi, M. Fiorini and M. Gabaldi, “Effects
of Annealing on Crystallinity and Phase Behavior of
PET/PC Block Copolymers,” European Polymer Journal,
Vol. 39, No. 6, June 2003, pp. 1081-1089.
[6] C. Carrot, S. Mbarek, M. Jaziri, Y. Chalamet, C. Raveyre
and F. Prochazka, “Immiscible Blends of PC and PET,
Current Knowledge and New Results: Rheological Prop-
erties,” Macromolecular Materials Engineering, Vol. 292,
No. 6, 2007, pp. 693-706. doi:10.1002/mame.200700006
[7] C. Carrot, S. Mbarek and M. Jaziri, “Recycling
Poly(Ethyelene Terephthalate) Wastes: Properties of Poly
(Ethylene Terephyhalate)/Polycarbonate Blends and the
Effect of a Transesterification Catalyst,” Polymer Engi-
neering & Science, Vol. 46, No. 10, 2006, pp. 1378- 1386.
[8] P. Marchese, A. Celi and M. Fiorini, “Influence of Activ-
ity of Transesterification Catalysts on the Phase Behavior
of PC-PET Blends,” Macromolecular Chemistry and
Physics, Vol. 203, No. 4, 2002, pp. 695-704.
[9] L. C. Mendes, A. M. Giornes, A. F. Cordeiro, M. R. Ben-
zi and M. L. Dias, “Miscibility of PET/PC Blends In-
duced by Cobalt Complexes,” International Journal of
Polymeric Materials, Vol. 56, No. 3, March 2007, pp.
257-272. doi:10.1080/00914030600812491
[10] P. S. C. Pereira, L. C. Mendes, L. Sirelli and M. L. Dias,
“Influence of Cobalt Complex on Thermal Properties of
Poly(Ethylene Terephthalate)/Polycarbonate Blend,” Jou-
rnal of Thermal Analysis and Calorimetry, Vol. 87, No. 3,
March 2007, pp. 667-671.
[11] P. S. C. Pereira, L. C. Mendes and R. E. R. Abrigo,
“Changes in Properties of PET/PC Blend by Catalyst and
Time,” International Journal of Polymeric Materials, Vol.
57, No. 4-6, 2008, pp. 494-505.
[12] L. C. Mendes, R. E. R Abrigo, V. D. Ramos and P. S. C.
Pereira, “Effect of Melt Flow Rate of Polycarbonate and
Cobalt Catalyst on Properties of PET/PC (80/20 wt%)
Reactive Blending,” Journal of Thermal Analysis and
Calorimetry, Vol. 99, No. 2, 2009, pp. 545-549.
[13] P. S. C. Pereira, L. C. Mendes and V. D. Ramos, “Rheo-
logical Study Bringing New Insights into PET/PC Reac-
tive Blends,” Macromolecular Symposia, Vol. 290, No. 1,
2010, pp. 121-131. doi:10.1002/masy.201050414
[14] W. Bu and J. He, “The Effect of Mixing Time on the
Morphology of Immiscible Polymer Blends,” Journal of
Applied Polymer Science, Vol. 62, No. 9, 1996, pp. 1445-
[15] P. Zahedi and A. Arefazar, “Blends of Poly(Ethylene
Terephthalate)/Polycarbonate by the Use of Lanthanum
Acetyl Acetonate Catalyst,” Journal of Applied Polymer
Science, Vol. 107, No. 5, 2008, pp. 2917-2922.
[16] G. Montaudo, F. Samperi and C. Puglisi, “Mechanism of
Exchange in PBT/PC and PET/PC Blends: Composition
of the Copolymer Formed in the Melt Mixing Process,”
Macromolecules, Vol. 31, No. 3, 1998, pp. 650-661.
Copyright © 2011 SciRes. MSA