Compatibilization and Property Characterization of

Polycarbonate/Polyurethane Polymeric Alloys

P. S. Archondouli, J. K. Kallitsis, N. K. Kalfoglou

Department of Chemistry, University of Patra, 26500 Patra, Greece

Received 11 October 2001; accepted 24 April 2002

Published online 11 February 2003 in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/app.11503

ABSTRACT: Polymeric alloys of bisphenol A polycarbon- previously with poly(ethylene terephthalate) or poly(buty-
ate and a polyester-type polyurethane were prepared over a lene terephthalate). This was attributed to the lower reactiv-
complete composition range through melt mixing. The ity of the aromatic OOH of polycarbonate toward the iso-
blends were characterized with optical and scanning elec- cyanate groups of polyurethane and the thermal instability
tron microscopy, tensile testing, dynamic mechanical anal- of the copolymer– compatibilizer formed. Experimental evi-
ysis, thermal analysis (differential scanning calorimetry and dence for the formation of the latter was provided by 1H-
thermogravimetric analysis), and spectroscopic techniques NMR and dynamic mechanical analysis of selectively
(Fourier transform infrared and 1H-NMR). A morphology leached blends and by quantitative analysis results for the
examination revealed good component dispersion with 30 extraction experiments. © 2003 Wiley Periodicals, Inc. J Appl
wt % polyurethane or less and strong interface adhesion. Polym Sci 88: 612– 626, 2003
Dynamic mechanical analysis indicated partial component
miscibility and tensile testing mechanical behavior typical of
a polymeric alloy. Overall, the degree of compatibilization Key words: polycarbonates; polyurethanes; alloys; compati-
obtained with polycarbonate was lower than that observed bilization; blends

INTRODUCTION TPUs play important roles within the rapidly grow-

ing family of thermoplastic elastomers because they
Polymer blending leading to a miscible system or a
possess exceptional properties, such as high tensile
polymeric alloy is a useful and more cost-effective
strength and elongation, flexibility, and low-tempera-
route than the synthesis of new polymers for the prep-
ture impact resistance. However, they have short-term
aration of materials with unique and tailor-made
resistance to elevated temperatures, and it would be
properties. preferable in some cases to increase their stiffness.
In recent work,1,2 the compatibility of a polyester- PC is a tough, transparent, impact-resistant thermo-
type thermoplastic polyurethane (TPU) and crystal- plastic engineering polymer with a very wide range of
line aromatic polyesters [poly(butylene terephthalate) service temperatures. However, it suffers from its high
(PBT) and poly(ethylene terephthalate) (PET)] was in- melt viscosity, which causes difficulties in processing
vestigated. The binary blends were compatible over (e.g., injection molding), and its low fatigue strength,
the complete composition range, and this was attrib- which generates cracks at points of concentrated re-
uted to an in situ PBT–polyurethane (PU) or PET–PU sidual stress.
copolymer formed via ester–amide interchange reac- For these reasons, blends of TPU and PC are ex-
tions during melt blending. pected to usefully combine the properties of the parent
In this article, we report on the compatibility behav- polymers.
ior of TPUs and bisphenol A polycarbonate (PC), an A short review of PC blends with polyesters is given
amorphous thermoplastic resin. in ref. 3. Other studies relevant to this one have dealt
with blends of aromatic polyesters and PU. Cyclohex-
ane cast films of PC/PU blends were stated to be
Correspondence to: N. K. Kalfoglou (n.kalfoglou@chemistry. incompatible. The PU was a polyether or polyester
upatras.gr). type.4 Partial compatibility with pyridine cast films
Contract grant sponsor: Programme on Polymer Science
and Technology-3.2a (33H6), Operational Programme for was reported by Ahn et al.5 The PU component was
Education and Initial Vocational Training, administered based on 4,4⬘-diphenyl methane diisocyanate (MDI)
jointly by the Greek Ministry of Education and the European and butanediol-1,4 (BD-1,4) as the hard segment and
Community (Programme JOULE 1998 –2000); contract grant poly(ethylene adipate) or poly(butylene adipate) as
number: PL-970014.
the soft segment. Films annealed in a differential scan-
Journal of Applied Polymer Science, Vol. 88, 612– 626 (2003) ning calorimeter between 230 and 270°C for various
© 2003 Wiley Periodicals, Inc. lengths of time showed a convergence of the main
CHARACTERIZATION OF PC/PU POLYMERIC ALLOYS 613

Figure 1 Phase-contrast micrographs of PC/PU blends: (a) 90/10, (b) 70/30, (c) 50/50, (d) 30/70, and (e) 10/90 (tmix ⫽ 5
min).

components’ glass-transition temperatures (Tg’s). This ysis (DMA), the system was stated to be miscible. The
was taken as evidence for reactive compatibilization miscibility was attributed to hydrogen-bonding inter-
attributed to ester–amide interchange reactions. No molecular interactions between ONH of the PU
other characterization techniques were applied. Melt groups and OCAO of the PC groups.
mixing at 235°C was employed by Fabri et al.6 to In this work, we examined PC/TPU blends pre-
obtain PC/PU elastomer blends. The examined com- pared by melt mixing over the complete composition
positions were limited to 0 –10 wt % PU. On the basis range. In addition to static and dynamic mechanical
of the Tg shift obtained with dynamic mechanical anal- testing, the degree of physical and reactive compati-
614 ARCHONDOULI, KALLITSIS, AND KALFOGLOU

bilization was investigated with spectroscopy [NMR

and Fourier transform infrared (FTIR)], optical micros-
copy (phase-contrast), and scanning electron micros-
copy (SEM) techniques.

EXPERIMENTAL
Materials and specimen preparation
PC was obtained from BDH Chemicals, Ltd., with a
number-average molecular weight (Mn) of 29,000 and
a density of 1.20 g cm⫺3. TPU (Desmopan 359) was
donated by Bayer A.G. It was a polyester-type PU
with hard segments formed by the addition of BD-1,4
to MDI. The soft segment, with Mn ⬃ 2000 g mol⫺1,
consisted of polyester chains formed by the polycon-
densation of adipic acid onto BD-1,4. The density was
1.23 g cm⫺3. The polymers were available in the form
of pellets. PC was dried at 120°C overnight for the
removal of sorbed water because even a low water
content could lead to hydrolysis of the polymer, par-
ticularly at elevated temperatures.7 PU was dried at
100°C for 24 h.
The blends were prepared via melt mixing in a
version of a mixer more automated than that de-
scribed previously.8
On the basis of the ultimate tensile properties, the
optimum mixing temperature and mixing time (tmix)
were determined to be 235°C and 5 min, respectively.
The compositions of the PC/PU blends prepared were
90/10, 70/30, 50/50, 30/70, and 10/90 (w/w).
The films were obtained via compression molding
between Teflon sheets at 235°C and 5 MPa, which was
followed by quenching to 0°C. They were conditioned
at approximately 25°C and at a relative humidity of
approximately 60% for 24 h.
Terminal hydroxyl (phenolic) groups in PC may be
important for the compatibilization process. So that
their presence could be established,9 PC was dissolved
in methylene chloride and precipitated in acetone. It
was filtered, redissolved, and reprecipitated, and this
process was repeated several times so that any resid-
ual contents of the bisphenol A and oligomers, both
soluble in acetone, would be removed. The second
stage involved the esterification of the purified PC Figure 2 Phase-contrast micrographs of 70/30 PC/PU
with an excess of acetyl chloride, in a solution of blends showing the effects of tmix: (a) 5, (b) 10, and (c) 15
min.
methylene chloride, by the presence of triethylamine.
Finally, it was precipitated in methanol, washed with
water, and dried at 60°C in vacuo overnight. acetyl end groups was in fairly good agreement with
A comparison of the NMR spectra of PC and its the expected number of terminal OH groups.
acetylated product, obtained from a solution in CDCl3,
indicated a new peak at 2.3 ppm in the latter spec-
Apparatus and procedures
trum. The literature10 confirms that this was due to the
methyl protons of the acetyl group. According to the Tensile tests were performed at a crosshead speed of
ratio of the new peak to the peak of the aromatic 10 cm min⫺1 and at 23°C, according to ASTM D 882,
protons or methyl protons of PC as well as its degree with a J.J. 5001 tensile tester and a film strip measuring
of polymerization (DP), the derived number for the 3.00 ⫻ 0.5 ⫻ 0.025 cm3.
CHARACTERIZATION OF PC/PU POLYMERIC ALLOYS 615

Figure 3 Phase-contrast micrographs showing the effects of annealing for 70/30 PC/PU blends [(a) quenched and (b)
annealed] and 30/70 PC/PU blends [(c) quenched and (d) annealed] and the effects of aging for 70/30 PC/PU blends [(e)
quenched and (f) aged] and 30/70 PC/PU blends [(g) quenched and (h) aged].

DMA data were obtained with a solid-state RSA II Differential scanning calorimetry (DSC) analyses
analyzer (Rheometric Scientific Co.) at 10 Hz. The were performed with a differential scanning calorim-
specimen dimensions were 3.0 ⫻ 0.5 ⫻ 0.01 cm3. eter (SP⫹) equipped with the Autocool accessory from
616 ARCHONDOULI, KALLITSIS, AND KALFOGLOU

heating rate was 10°C min⫺1. The samples were

heated from 25 to 250°C.
To examine the effect of tmix on possible decomposi-
tion, we performed thermogravimetric analysis (TGA)
isothermal scans under a constant flow of dry nitrogen
with a DuPont 951 thermogravimetric analyzer
equipped with a 990 program recorder. The samples
were tested at 235°C for 20 min.
Optical micrographs with a phase-contrast arrange-
ment were obtained with an Olympus BH-2 micro-
scope.
SEM was carried out on a JEOL JSM-500 instrument
at a tilt angle of 0° on specimens fractured at cryogenic
and ambient temperatures.
Extraction experiments were performed with chlo-
roform for 5 days at ambient temperature. In this
solvent, PC was completely soluble, but PU remained
practically undissolved (amount dissolved ⬃ 1 wt %).
After extraction, the films were dried in a vacuum
oven at 60°C for 3 days.
FTIR spectra were recorded with a PerkinElmer
1600 spectrometer.
1
H-NMR spectroscopy was performed with an
Avance DPX 400-MHz instrument with a magnetic
field strength of 9.4 T. The samples were dissolved in
a 2/1 (v/v) mixture of CF3COOH and CDCl3 so that 2
wt % solutions would be obtained.

RESULTS
Morphology
Optical microscopy
Given the refractive-index difference of PU (nD ⫽
1.550)11 and PC (nD ⫽ 1.584),7 positive phase-contrast
dark areas would correspond to PC.
Figure 1 shows the effect of the composition varia-
tion on blends quenched at 25°C. Good component
dispersion was evident for the 90/10 blend [Fig. 1(a)].
Increasing the PU content led to a phase-separated
system [Fig. 1(b– e)]. Figure 1(b,c) shows a character-
istic interpenetrating morphology, but at PU compo-
sitions greater than 50 wt %, matrix inversion took
place. As in previous polyester/PU blends,1,2 PC sep-
arated as spherical domains that were composite in
nature; that is, PU was occluded in the PC globules.
Note the light gray areas inside the darker PC glob-
ules. A larger contrast was observed in the previous
blends1,2 because of the larger refractive-index differ-
ence between components. The effect of tmix is shown
in Figure 2. Some improvement in the dispersion was
Figure 4 SEM images of cryofractured surfaces of PC/PU observed when tmix was increased to 10 min. How-
blends: (a) 90/10, (b) 70/30, (c) 50/50, and (d) 30/70.
ever, the interpenetrating nature of the blend mor-
phology persisted. On a macroscale, an improved dis-
Rheometric Scientific. All the experiments were per- persion was observed after the annealing of the 70/30
formed under a constant flow of dry nitrogen. The and 30/70 blends, which led to a pronounced spheri-
nominal weight was approximately 10 mg, and the cal domain morphology for the latter [cf. Fig. 3(a– d)].
CHARACTERIZATION OF PC/PU POLYMERIC ALLOYS 617

Figure 5 SEM images of cryofractured surfaces of PC/PU blends at a higher magnification: (a) 90/10, (b) 70/30, (c) 50/50,
and (d,e) 30/70.
618 ARCHONDOULI, KALLITSIS, AND KALFOGLOU

Figure 6 SEM images of PC/PU blends fractured at ambient temperature: (a,b) 90/10 and (c,d) 70/30.

Aging had no effect on the morphology, the high elongation, the dewetting of the PC particles at their
component viscosity preventing any morphological equator took place [see the gaps around the inclusions
transformation [cf. Fig. 3(e– h)]. in Fig. 6(c)].

Mechanical properties
SEM
Tensile properties
SEM micrographs in Figures 4 – 6 give additional in-
formation on the phase distribution and component The data and standard deviations for the ultimate
adhesion of the fractured surfaces. In Figure 4, the properties, including the yield stress (␴y), are summa-
transition from a morphology with a fracture typical rized in Table I for blends quenched to 0°C. A large
of a homogeneously mixed blend [Fig. 4(a)] to an deformation behavior in terms of the tensile strength
interpenetrating morphology [Fig. 4 (b,c]] to a spher- (␴b) and, in particular, the elongation at break (⑀b), a
ical domain morphology [Fig. 4(d)] is evident. At a sensitive indicator of component adhesion in blends,
higher magnification (Fig. 5), these morphologies are was used to evaluate the mechanical performance and
amplified, and in Figure 5(b– e), one can discern the to optimize the mixing conditions. Mixing at temper-
fissures caused by the expansivity mismatch of the atures greater than or equal to 260°C resulted in the
two components at cryogenic temperatures. For the deterioration of the mechanical properties, and this, as
30/70 composition, quite a few of the PC globules well as the effect of annealing, was discussed in the
were firmly embedded in the PC matrix. In Figure 6, last section. Annealing led to an enhanced phase sep-
fracturing at the ambient temperature shows a mixed aration, which was confirmed by DMA and phase-
glassy– ductile morphology for the 90/10 composi- contrast microscopy. The lower tmix value was
tion, the rubbery component being elongated [Fig. adopted because pertinent data showed less scatter
6(a)] but anchored in the glassy matrix [Fig. 6(b)]. In and greater tmix values led to higher weight losses in
Figure 6(c,d), one can observe that the composite glob- TGA (discussed later). Aging did have an adverse
ules of PC, although broken during deformation, re- effect on the tensile properties of a PC-rich blend
mained attached to the PU matrix. Because of matrix (70/30), possibly the result of phase agglomeration,
CHARACTERIZATION OF PC/PU POLYMERIC ALLOYS 619

TABLE I
Ultimate Tensile Properties of Quenched Blendsa
PC/PU Mixing ␴y ␴b ␧b Eb
blend temperature (°C) (MPa) (MPa) (%) (J/cm3)

100/0 — 54 ⫾ 2 63 ⫾ 3 167 ⫾ 16 89 ⫾ 10
90/10 235 47 ⫾ 1 55 ⫾ 7 182 ⫾ 19 82 ⫾ 13
70/30 235 34 ⫾ 3 49 ⫾ 8 200 ⫾ 32 80 ⫾ 12
70/30 240 32 ⫾ 1 43 ⫾ 6 209 ⫾ 77 84 ⫾ 22
70/30 270 32 ⫾ 2 33 ⫾ 2 41 ⫾ 11 14 ⫾ 6
50/50 235 — 33 ⫾ 3 171 ⫾ 15 46 ⫾ 4
30/70 235 — 25 ⫾ 3 243 ⫾ 31 39 ⫾ 6
10/90 235 — 26 ⫾ 3 319 ⫾ 46 37 ⫾ 16
0/100 — — 56 ⫾ 5 807 ⫾ 130 222 ⫾ 22
70/30b 235 34 ⫾ 7 36 ⫾ 7 60 ⫾ 18 18 ⫾ 8
30/70b 235 — 18 ⫾ 2 130 ⫾ 44 20 ⫾ 3
70/30c 235 33 ⫾ 3 35 ⫾ 3 101 ⫾ 24 30 ⫾ 7
30/70c 235 — 26 ⫾ 2 213 ⫾ 52 37 ⫾ 10
70/30d 235 33 ⫾ 2 43 ⫾ 4 228 ⫾ 73 92 ⫾ 33
70/30c 235 33 ⫾ 2 49 ⫾ 6 211 ⫾ 40 79 ⫾ 8
a
tmix ⫽ 5 min.
b
Annealed at 150°C for 1 h.
c
After 40 days of physical aging at 25°C.
d
tmix ⫽ 10 min.
e
tmix ⫽ 15 min.

although microscopy findings did not reveal it. An- the composition dependence of this useful quantity. Eb
other possibility is cited later (see the DSC results). In (or the impact strength) was maintained at a level
Table I, Eb represents the energy to tensile failure, and comparable to that of PC up to approximately 30 wt %
it was determined from the area enclosed by the PU.
stress–strain curve. It is related to the impact strength, In general, the ultimate properties were character-
and for a series of similarly prepared blends, it gives istic of a polymeric alloy with good component inter-

TABLE II
Viscoelastic and Thermal Properties of Quenched Blends
a
PC/PU High Tg Low Tg Tm of PUh Tm of PC Xc of PU
blend (°C) (°C) (°C) (°C) (%)

100/0b 153 — — 240 —

100/0 155 — — — —
90/10 148 — — — —
70/30 141 25 200 — 10
50/50 147 30 202 — 21
30/70 153 16 175,187,200 — 16
10/90 153 12 177, 188, 203 — 21
0/100 — 3 176, 203 — 22
70/30c 148 30 199 — 4
70/30d 145 28 — — —
70/30e 149 ⫺20 201 — 18
30/70e 153 ⫺20 203, 218 — 17
70/30f NE NE 200 — 11
30/70f NE NE 175, 187, 202 — 21
70/30g 132 3 — 235 i
30/70h — 18 202 220 21

NE ⫽ not examined.
a
⌬H0f of (PU) ⫽ 5.83 cal/gr.
b
After treatment with CHCl3.
c
tm ⫽ 10 min.
d
tm ⫽ 15 min.
e
After annealing at 150°C for 1 h.
f
After 40 days of physical aging at 25°C.
g
Soluble fraction.
h
Insoluble fraction.
620 ARCHONDOULI, KALLITSIS, AND KALFOGLOU

Figure 7 Temperature dependence of E⬘ for PC/PU blends: () 100/0, () 90/10, () 70/30, () 50/50, () 30/70, (〫) 10/90, and
() 0/100. The inset shows E⬘ for leached blends: (Œ) 70/30, soluble fraction, and () 30/70, insoluble fraction.

face adhesion (see the ⑀b percentage data). The drop at 50 wt % PU, the polycarbonate glass-transition tem-
the median compositions was expected for nonmis- perature (Tg,PC) reversed its trend, approaching the Tg
cible blends and was attributed to matrix inversion. value of the pure component. This was corroborated
by the morphology findings (va), which revealed
phase-separated PC domains (see Fig. 1). The insets in
Dynamic mechanical properties
Figures 7 and 8 refer to the spectra of blends in which
DMA data for blends may provide information about one of the components was removed. The Tg shifts for
the degree of polymer–polymer interactions and inter- both components persisted (see also Table II). This is
phase mixing. relevant for explaining the compatibilization process
The main relaxations (Tg) of the blends are reported (see the Discussion section).
for various compositions in Table II. Figures 7 and 8 Increasing tmix beyond 5 min did not lead to a
give the DMA spectra under isochronous conditions further Tg,PC decrease. Apparently, the degree of dis-
for the storage modulus (E⬘) and loss modulus (E⬙), persion on the Tg scale was not improved.
respectively, for quenched blends at tmix ⫽ 5 min. Both Annealing mainly reduced Tg,PU, and this was pos-
primary relaxations are present, although they con- sibly the effect of enhanced crystalline–amorphous
verge toward each other as the composition changes. phase separation of the PU component (see Table II
Overall, the spectra are typical of a phase-separated and Figure 9). It is expected that when crystalline
polymeric alloy. In general, the shift in the polyure- blocks are closely intermingled with amorphous seg-
thane glass-transition temperature (Tg,PU) is more sig- ments, the Tg value of the amorphous phase will be
nificant than the shift for the polyester. The addition of raised. This was the case for the quenched specimens
PU flexibilized PC at PU concentrations greater than for which Tg was approximately 3°C (see Table II).
10 wt % (see Fig. 8). At compositions with greater than This effect of the thermal history on Tg was also re-
CHARACTERIZATION OF PC/PU POLYMERIC ALLOYS 621

Figure 8 Temperature dependence of E⬙ for PC/PU blends: () 100/0, () 90/10, () 70/30, () 50/50, () 30/70, (〫) 10/90, and
() 0/100. The inset shows E⬙ for leached blends: (Œ) 70/30, soluble fraction, and () 30/70, insoluble fraction.

ported for PU12 and poly(ether ester) copolymers that crystallite perfection was less readily attained, endo-
also had short crystalline–amorphous sequences along therm multiplicity being the rule. Physical aging did
their main chains.13 not affect XC of a PC-rich blend (70/30), and so its
effect on mechanical properties (va) could be attrib-
uted to an amorphous-phase densification. The effect
Thermal properties
of tmix in suppressing crystallinity was significant for
DSC results the 70/30 blends; no XC and Tm values were detected
after tmix ⫽ 15 min. The fraction of the blend soluble in
The DSC results are summarized in Table II and were
CHCl3 gave a Tm,PC value caused by solvent-induced
obtained during the first heating scan. PU crystallized
crystallization.7 For the 30/70 blends, the phase sepa-
slowly, giving a multiplicity of endotherms depend-
ration of PC allowed crystallinity to be developed at a
ing on the thermal history and rate of heating, the
level comparable to that of PU. In general, the multi-
highest melting temperature (Tm ⫽ 203°C) corre-
plicity and overlapping of endotherms did not allow a
sponding to the most perfect crystal. At a high dilu-
very accurate determination of XC of PU. Therefore,
tion (90/10) in a medium (PC) of high viscosity, crys-
the importance of the XC values cannot be overempha-
tal formation was suppressed, so no Tm and no bulk
sized.
crystallinity (XC) were detected. In compositions in
which the phases were separated, crystallinity began
to develop, and Tm was slightly depressed; this at-
TGA results
tested to a lesser degree of crystal perfection (70/30
blends). At higher PU contents, when matrix inversion Figure 10 summarizes the results of TGA for the 70/30
took place and PC composite domains were formed, blends (tmix ⫽ 5 or 10 min) under isothermal condi-
crystallinity development was less hindered, but PU tions (235°C).
622 ARCHONDOULI, KALLITSIS, AND KALFOGLOU

Figure 9 Effect of annealing on E⬙ for PC/PU blends: () 70/30, quenched; (‚) 70/30, annealed; () 30/70, quenched; and (ƒ)
30/70, annealed. The inset shows E⬙.

PC remained practically stable at this temperature, selectively leaching out one blend component with
but there was significant chain scission in PU, as CHCl3. The spectra of the nonleached blends did not
shown by the observed weight loss, which, in princi- show any additional features or shifts. The spectra of
ple, could lead to reactive compatibilization. More- the leached blends were expected to reveal possibly
over, the 70/30 blends with tmix ⫽ 5 or 10 min showed the incorporation of the soluble partner (PC) into the
considerably higher weight losses. This indicated that insoluble PU or any chemically formed species (see
the materials obtained after blending behaved differ- Fig. 11).
ently than expected for a physical mixture of the two A comparison of the PU and PC spectra with those
components, supporting the view that interchain re- of the leached blends (70/30 and 30/70, insoluble
actions occurred during the blending process (see the fractions) indicated that in both compositions, al-
last section). The higher PU loss in the blends in though much of the PC was removed, there were still
comparison with pure PU was previously invoked5 as peaks at 1.73 (methyl protons) and 7.30 ppm (aromatic
evidence for reactive compatibilization. protons), which were characteristic of its presence.
A quantitative analysis of these spectra showed that
Spectroscopy for the PU-rich blend (30/70), the insoluble residue
1 after extraction had a 20/80 composition, whereas for
H-NMR the 70/30 blend, the remaining residue had a 50/50
To obtain evidence for possible trans reactions be- composition. A similar examination of the soluble
tween PC and PU during blending, we compared fraction of the 30/70 blend revealed that a small
1
H-NMR spectra of the pure components with the amount of PU was solubilized together with PC and in
spectra of the 70/30 and 30/70 blends before and after excess of the unmodified PU (ca. 1 wt %) determined
CHARACTERIZATION OF PC/PU POLYMERIC ALLOYS 623

Figure 10 TGA results for (—) PC, tmix ⫽ 5 min; (–) PU, tmix ⫽ 5 min; (. . .) 70/30 PC/PU, tmix ⫽ 5 min; and (- 䡠 -) 70/30
PC/PU, tmix ⫽ 10 min (isothermal conditions, 235°C).

by separate experiments. This implied the formation a chemical modification of the PU chain during mixing
of a PC–PU copolymer corresponding to approxi- or the existence of traces of PU or modified PU in the
mately 3 wt % PU (based on the total amount of the extract (va).
extract).

DISCUSSION
FTIR
The morphology findings and large deformation me-
FTIR spectra of blends with various compositions did chanical behavior support the view that melt-mixed
not reveal any new absorptions or any differentiation PC/PU blends form polymeric alloys over the com-
in the spectral region associated with intramolecular
plete composition range. At high PC contents (90/10),
hydrogen bonding at 3332 cm⫺1 due to the PU imide–
both optical microscopy and SEM revealed a high
ester group interaction.1 Therefore, the proposition
degree of dispersion. Increased amounts of PU (ⱖ50
that compatibilization is attained via hydrogen bond-
wt %) gave composite inclusions of PC, as observed
ing between urethane and carbonate groups is not
substantiated by this technique because this intermo- previously.1,2
lecular bonding would also alter the spectrum in the In our previous work,1,2 conclusive evidence was
aforementioned region. presented to support the view that PET and PBT dur-
As before,1 to obtain evidence for possible reactive ing melt mixing were reactively compatibilized with
compatibilization via copolymer formation, we freed TPUs. Both FTIR and 1H-NMR gave evidence for co-
the 70/30 and 30/70 blends from PU by prolonged polymer formation, the result of ⬃NCO groups react-
extraction with CHCl3 to obtain soluble PC in the ing with the aliphatic terminal OOH of PET or PBT. In
extract. In Figure 12, the absorption of the OCAO this system, the available OOH of PC was aromatic;
group of PC films in the range 1760 –1790 cm⫺1 is its reactivity with the terminal ⬃NCO groups (ob-
shown to be altered and seems to be coupled with the tained by scission of the PU chain) was lower than that
amide I conformation of OCAO and ONH groups14 with the aliphatic OOH.15 In addition, the copolymer
of PU in the range 1690 –1745 cm⫺1. This may support that formed according to the following reaction
 624

Figure 11 NMR spectra of leached PC/PU blends: (a) pure PC; (b) pure PU; (c) 70/30 PC/PU, insoluble fraction; and (d) 30/70
ARCHONDOULI, KALLITSIS, AND KALFOGLOU

PC/PU, insoluble fraction.

CHARACTERIZATION OF PC/PU POLYMERIC ALLOYS 625

Figure 12 FTIR spectra of the leached PC/PU blends (soluble fractions).

1. 1H-NMR spectra of leached blends showed the

presence of the leached-out component (PC) in
the insoluble fractions as well as the presence of
some modified PU (possibly as a copolymer) in
(1) the extract (va).
2. These findings agreed with selective extraction
experiments with CHCl3 for the removal of any
free (not bonded to PU) PC from the blends.
Therefore, in the 30/70 blends, the insoluble res-
was stated to be unstable.16 Therefore, the low reac-
idue had a composition of 14/86; 1H-NMR gave
tivity and thermal instability reduced the amount of
a 20/80 ratio. Similarly, for the 70/30 blends, the
the ester–amide copolymer that could form. More-
residue after extraction had a composition of ap-
over, as Table I indicates, an increased temperature of
proximately 60/40 versus 50/50 from 1H-NMR.
mixing lowered drastically the tensile properties of the
blends. Annealing also had an adverse effect on these One cannot attach a high degree of accuracy to
properties and promoted phase separation, as DMA these determinations because a small percentage
and optical microscopy indicated. In addition, if reac- of PU homopolymer and some PC–PU copoly-
tive compatibilization proceeded via reaction (1), as mer were also leached with the unattached PC
shown in our previous study,1 no modification of the (va).
spectrum by the aromatic protons was expected be- 3. More evidence was provided by DMA, which
cause no change in the environment took place on indicated that the main component relaxations
account of the formation of the PC–PU copolymer. converged and that these shifts were maintained
However, there was other evidence that suggested even after the extraction and removal of one com-
that a copolymer might form: ponent (PC; see Table II). The small amount of
626 ARCHONDOULI, KALLITSIS, AND KALFOGLOU

PU that might have been mixed with PC in the The authors thank K. Gravalos and P. Tsolakis for useful
extract could not account for the Tg,b shifts ob- discussions. Thanks are also due to Bayer A.G. for supplying
served. Compatibilization based on intermolecu- the polyurethane.
lar forces (e.g., hydrogen bonding) could not be
adopted because the evidence from FTIR did not
substantiate it. Therefore, some chemical modifi- References
cation of the blend partners was involved.
4. As mentioned previously, TGA also provided 1. Archondouli, P. S.; Kalfoglou, N. K. Polymer 2001, 42, 3489.
indirect evidence for enhanced PU reactivity in 2. Samios, C. K.; Gravalos, K. G.; Kalfoglou, N. K. Eur Polym J
2000, 36, 937.
the presence of PC (va).
3. Samios, C. K.; Kalfoglou, N. K. Polymer 2000, 41, 5759.
4. Peterson, R. J.; Corneliussen, R. D.; Rozelle, L. T. Polym Prepr
The other possibility for PC reacting with PU via a (Am Chem Soc Div Polym Chem) 1969, 10, 385.
transesterification reaction was not substantiated by 5. Ahn, T. O.; Jung, S.; Lee, J.; Jeong, H. M. J Appl Polym Sci 1997,
FTIR. Had this been the case, the absorbance ratio of 64, 2363.
the peaks at 1780 and 1760 cm⫺1 would have been 6. Fabri, L.; Penati, A.; Kolarik, J. Polymer 1997, 38, 835.
altered.17 This was not observed. However, this is only 7. Freitag, D.; Grigo, U.; Müller, P. R.; Nouvertné, W. In Encyclo-
pedia of Polymer Science and Engineering, 2nd ed.; Mark, H. F.;
a tentative conclusion, and further work is needed to
Bikales, N. M.; Overberger, C. G.; Menges, G., Eds.; Wiley-
arrive at a definite statement. Interscience: New York, 1989; Vol. 11, p 648.
8. Papadopoulou, C. P.; Kalfoglou, N. K. Polymer 1997, 38, 4207.
9. Horbach, A.; Veiel, U.; Wunderlich, H. Makromol Chem 1965,
CONCLUSIONS
88, 215.
PC/PU blends were compatibilized by melt mixing 10. Hesse, M.; Meier, H.; Zeeh, B. Spectroscopic Methods in Organic
via a route similar to that described previously,1,2 that Chemistry; Thieme: Stuttgart, 1997; Chapter 3.
11. Seymour, R. B.; Carraher, C. E. Structure–Property Relation-
is, the formation of a copolymer combining polyester ships in Polymers; Plenum: New York, 1984; Chapter 4.
and PU segments. For reasons cited previously, the 12. Abouzahr, S.; Wilkes, L. J Appl Polym Sci 1984, 29, 2711.
concentration of this copolymer was quite low. There- 13. Kalfoglou, N. K. J Appl Polym Sci 1984, 29, 2711.
fore, a lower degree of compatibilization was attained 14. Bellamy, L. J. The Infra-Red Spectra of Complex Molecules, 2nd
for the PC/PU blends in comparison with the PET/PU ed.; Methuen: London, 1958; Chapter 12.
and PBT/PU blends. Although formed at low levels, 15. Ravve, A. Organic Chemistry of Macromolecules; Marcel Dek-
ker: New York, 1967; p 295.
this copolymer could be effective as a compatibilizer
16. Wood, G. The ICI Polyurethane Book, 2nd ed.; Wiley: New
by reducing the interfacial tension of the blend com- York, 1990; Chapter 3.
ponents, thereby facilitating their dispersion and pro- 17. Devaux, J.; Godard, P.; Mercier, J. P.; Touillaux, R.; Dereppe,
moting the compatibilization process. J. M. J Polym Sci Part B: Polym Phys 1982, 20, 1881.