Blends of Poly(ethylene terephthalate) and Poly(butylene

terephthalate)

G. Aravinthan, D. D. Kale
Polymer Engineering and Technology Department, University of Mumbai, Institute of Chemical Technology,
Matunga, Mumbai 400 019, India

Received 21 May 2004; accepted 27 December 2004

DOI 10.1002/app.22017
Published online in Wiley InterScience (www.interscience.wiley.com).

ABSTRACT: Commercial grade poly(ethylene terephtha- sition. The addition of PBT increased the processability of
late), (PET, intrinsic viscosity ⫽ 0.80 dL/g) and poly(buty- PET. Differential scanning calorimetry data showed the
lene terephthalate), (PBT, intrinsic viscosity ⫽ 1.00 dL/g) presence of both phases. For all blends, only a single glass-
were melt blended over the entire composition range using transition temperature was observed. The melting character-
a counterrotating twin-screw extruder. The mechanical, istics of one phase were influenced by the presence of the
thermal, electrical, and rheological properties of the blends other. © 2005 Wiley Periodicals, Inc. J Appl Polym Sci 98: 75– 82,
were studied. All of the blends showed higher impact prop- 2005
erties than that of PET or PBT. The 50:50 blend composition
exhibited the highest impact value. Other mechanical prop- Key words: blends; poly(ethylene terephthalate); poly(buty-
erties also showed similar trends for blends of this compo- lene terephthalate); rheology

INTRODUCTION scanning calorimetry (DSC), IR spectroscopy, and

wide-angle X-ray diffraction (WAXD) techniques.
Poly(ethylene terephthalate) (PET) and poly(butylene
DSC studies showed that PBT and PET crystallized
terephthalate) (PBT) are both very important commer-
separately in the melt blends and one phase enhanced
cial aromatic–aliphatic polyester resins. PET is widely
the crystallization process of the other. IR spectros-
used for the fabrication of molded articles, fibers, and
copy indicated a separate crystallization rate for each
films. PBT is mainly used for injection molding and
component in the blends by showing the change in
related applications. PET resins are utilized where
intensity of the 917 and 848 cm⫺1 bands. WAXD stud-
higher a heat-deflection temperature, greater rigidity,
ies of PET/PBT melt blends showed diffraction char-
and economy are desired. However, the processing of
acteristics of the virgin homopolymers. No apprecia-
PET is difficult because of its low melt strength and
ble shifts in peak positions or new diffractions were
slow rate of crystallization. PBT resins have the ad-
observed. Avramova7 reported a single glass-transi-
vantages of rapid crystallization rate and good mold- tion temperature (Tg), indicating the miscibility of PET
ability. In general, the mechanical properties as well as and PBT segments in the amorphous phase. PET/PBT
chemical resistance characteristics are comparable for blends exhibited only one crystallization peak but sep-
both PET and PBT. A major drawback of PET and PBT arate melting points for the two crystalline species.
resins is their brittleness.1–3 Some depression of the melting point of one compo-
Mishra and Deopura4 predicted the theoretical com- nent associated with the presence of the other compo-
patibility of PET and PBT blends through thermody- nent indicated compatibility in the amorphous phase.
namic considerations. In another study,5 they argued Avramova7 also reported the presence of spherulite
that, although PET is compatible with PBT in the crystallization of both species in PET/PBT blends with
amorphous phase, it forms separate crystals rather microscopic observation. He concluded that the minor
than cocrystals. Therefore, these blends can give sig- component is excluded within the spherulites of the
nificantly different properties than PET or PBT, de- major component and may subsequently crystallize in
pending upon the composition and processing condi- this location. Jin and Tao8 reported the rheological
tions. Stein and coworkers6 studied the miscibility and properties of PET/PBT blends at a temperature of
morphology of PET/PBT blends using differential 275°C and a shear rate up to 103 s⫺1. They observed
well-defined zero shear viscosity for all blends and
shear thinning behavior in the high shear rate region
Correspondence to: D. D. Kale ([email protected]).
and a network type of morphology for a lower con-
Journal of Applied Polymer Science, Vol. 98, 75– 82 (2005) centration of PBT. They argued that the blend contain-
© 2005 Wiley Periodicals, Inc. ing 50% PBT was not suitable for fiber spinning be-
76 ARAVINTHAN AND KALE

cause of a sudden loss of elastic effect. This suggests head speeds were 50 and 2.8 mm/min for tensile tests
miscibility of PET and PBT over the entire composi- (ASTM 638M-91) and flexural tests (ASTM 790M-92),
tion range. A fascinating feature for blends in the respectively.
polyester family is the potential for transreaction, The notch for impact testing was cut using a motor-
which is also called rereaction or transesterification. ized notch-cutting machine (Polytest model 1, Ray
Porter and Wang9 suggested that miscibility in PET/ Ran). The notched Izod impact strength was measured
PBT blends is not a consequence of transesterification, at ambient conditions according to the ASTM D 256
when mixing times are short as in extrusion. It is method using an impact tester (Avery Dension) with a
known that melt mixing over a long period of time at striking velocity of 3.46 m/s and employing a 2.7-J
high temperature can lead to a transesterified random striker.
copolymer.
Several patents10 –12 mention good properties of
PET/PBT blend compositions. However, they do not Electrical properties
disclose actual values for these properties. There are The breakdown voltage for PET/PBT blends was mea-
no data in the literature on the systematic variation of sured using an AC High Voltage Tester (Zaran, Mum-
properties with composition for the important PET/ bai, India).
PBT blends. Therefore, the present study is useful.

Thermal properties
EXPERIMENTAL
Heat distortion temperature (HDT)
Materials
Specimens similar to those used for measuring the
Commercial grades of PET (RELPET, grade G5801,
flexural strength were used to determine the HDT
intrinsic viscosity ⫽ 0.80 dL/g) and PBT (LUPOX,
according to ASTM D 648. An HDT Vicat softening
GP-1000, intrinsic viscosity ⫽ 1.00 dL/g) were pro-
point apparatus (Davenport) was employed for mea-
cured from Reliance Industries Ltd. (Mumbai, India)
surements.
and LG Polymers India Pvt Ltd., respectively.

DSC
Blending
The melting and crystallization behavior of PET/PBT
PET and PBT chips were manually mixed and dried at
blends was studied using a DSC-7 (Perkin–Elmer).
120°C for 8 h in an air circulating oven. The blend
Temperature calibration was performed using indium
compositions were 80:20, 70:30, 60:40, 50:50, 40:60, and
as a reference (melting temperature ⫽ 156.60°C, heat
20:80 in which the numbers represent the weight per-
flow ⫽ 28.5 J/g). The heating rate of the samples was
centages of PET and PBT, respectively. A blend was
20°C/min with a sample size between 7 and 13 mg
also prepared by blending 60 parts PET with 40 parts
using a standard aluminum sample pan. The sample
PET/PBT (50:50), so that the final composition was 80
was maintained under a nitrogen atmosphere.
parts PET and 20 parts PBT. This blend is designated
as 80:20 A. The mixture was melt blended using a
counterrotating twin-screw extruder (Haake Rheocord Rheological properties
9000, TW100). The screw speed was maintained at 40
rpm and the temperature profiles for blending were The melt rheology of the individual polymers and
150°C for zone 1, 220°C for zone 2, 240°C for zone 3, their blends was studied using a rotational rheometer
and 260°C for the die zone. The extrudates were water (Haake RT 10), employing a parallel plate sensor with
cooled and pelletized. The extruded pellets were dried a diameter of 35 mm at 260°C. The shear rate range
and molded into standard ASTM test specimens by was varied from 1 to 10 s⫺1. For higher shear rates,
using a microprocessor-based injection-molding ma- rheological measurements at 260°C were carried out
chine (Boolani Industries Ltd., Mumbai, India). The using a twin-bore capillary rheometer (Rosand Preci-
temperature profiles were 220°C for zone 1, 250°C for sion). The shear rate range was 100 –1000 s⫺1. The
zone 2, and 280°C for the nozzle zone. capillary rheological data were subjected to Bagley
and Rabinowitch corrections.

Mechanical properties
Microscopy
Tensile strength, flexural strength, and flexural mod-
ulus were measured at ambient conditions using a The morphological characteristics of PET/PBT blends
universal tensile testing machine (LR-50K, Lloyds In- were examined by scanning electron microscopy with
strument), according to ASTM procedures. The cross- a Shimadzu SS-550 Superscan.
BLENDS OF PET AND PBT 77

TABLE I
Mechanical and Electrical Properties of PET/PBT Blends
Tensile Flexural Flexural Impact Breakdown
Compositions strength strength modulus strength HDT voltage
(wt %) (MPa) (MPa) (MPa) (J/m) (°C) (kV)

PET 56 76 2105 25 78 27
PET/PBT
80:20 53 75 2305 38 72 22
80:20 A 54 67 2872 41 75 22
70:30 51 70 2151 36 68 20
60:40 58 78 2287 44 67 24
50:50 62 73 2457 52 65 26
40:60 59 76 2321 42 60 20
20:80 57 73 2345 40 59 16
PBT 55 75 2127 27 56 14

RESULTS AND DISCUSSION strength. The impact strength as well as the flexural
modulus for the 50:50 PET/PBT blend are appreciably
Mechanical properties
higher than the virgin polymers. It is interesting to
Table I shows the variation of the mechanical proper- note that the impact strength of all the blends is higher
ties, electrical properties, and HDT for PET/PBT than that of virgin polymers. Because a blend of com-
blends. The tensile strength and flexural strength of position 50:50 PET/PBT exhibited the highest impact
PET and PBT are comparable. Therefore, it is not strength, a blend was prepared by blending 60 parts
surprising that both of these properties for all blends PET with 40 parts 50:50 PET/PBT blends so that the
are relatively comparable, although the 50:50 PET/ final composition of the blend was 80 parts PET and 20
PBT blend tends to show slightly higher tensile parts PBT. The mechanical properties of this blend are

Figure 1 DSC thermograms of PET/PBT blends in the heating cycle: (Œ) PET, (E) PET 80, (〫) PET 60, (⫻) PET 50, (F) PET
20, and (‚) PBT.
78 ARAVINTHAN AND KALE

Figure 2 DSC thermograms of PET/PBT blends in the cooling cycle: (Œ) PET, (E) PET 80, (〫) PET 60, (⫻) PET 50, (F) PET
20, and (‚) PBT.

also reported in Table I. Note that the impact strength conditions and miscibility, as suggested by Mishra
and flexural modulus of these blends increased and Deopura.5
slightly compared to the blends prepared by blending
80 parts PET and 20 parts PBT. Thus, it is clear that the
HDT analysis
50:50 PET/PBT blend composition shows better prop-
erties than virgin PET and PBT. The “interesting” The HDT of PET is much higher than that of PBT. As
properties of 50:50 PET/PBT blends are mentioned in the amount of PET in the PET/PBT blends increased,
the patent literature.10 The positive deviation of the the HDT steadily increased. Thus, the addition of PET
impact values may be correlated with the processing into PBT helps enhance the HDT of PBT. The blend

TABLE II
Thermal Characterization Data for PET/PBT Blends
Melting
Crystallization Heat of
temperature (°C) Heat of fusion (J/g)
Compositions temperature crystallization Tg
(wt %) PET PBT PET PBT (°C) (J/g) (°C)

PET 256 — 34 — 166 21 81

PET/PBT
80:20 255 — 35 (27) — (10) 193 33 75
80:20 A 251 — 47 — 189 38 71
70:30 253 221 26 (24) 4 (16) 187 34 73
60:40 253 222 24 (20) 15 (21) 186 36 72
50:50 252 225 17 (17) 13 (13) 183 40 69
40:60 253 227 15 (14) 21 (31) 185 40 67
20:80 254 233 7 (7) 36 (42) 188 43 64
PBT — 233 — 52 192 51 58
BLENDS OF PET AND PBT 79

Figure 3 The shear viscosity behavior of PET/PBT blends at 260°C: (Œ) PET, (E) PET 80, (⫹) 80:20 A PET/PBT, (䊐) PET 70,
(〫) PET 60, (⫻) PET 50, (■) PET 40, (F) PET 20, and (‚) PBT.

prepared by mixing 60 parts PET with 40 parts 50:50 PET/PBT blend composition, show two melting peaks
PET/PBT blends seems to show slightly higher HDT corresponding PBT and PET phases. The melting peak
compared to a normal blend of 80 parts PET and 20 temperature for PET is lower in the blends in compar-
parts PBT. ison to virgin PET. The peak melting temperature for
50:50 PET/PBT blends is the lowest. Similarly, the
Electrical properties peak melting temperature of PBT is also lower in
blends compared to virgin PBT. The lowering of this
Table I also shows the variation of the breakdown temperature for PBT appears to be related to the sys-
voltage for PET/PBT blends. Virgin PET has a higher tematic variation in the compositions.
breakdown voltage than virgin PBT. The addition of The heat of fusion (␦Hm) values show very interest-
PBT into PET in the PET/PBT blend compositions ing behavior. The ␦Hm value for PET-rich blends is
decreased the breakdown voltage. However, 50:50 higher than the value expected from its amount
PET/PBT blends showed appreciably higher values present in the blends based upon the linear additivity
than other compositions. rule. The values in the parentheses in Table II corre-
spond to the additivity rule. Similarly, the ␦Hm values
Thermal properties for PBT are also shown in parentheses. It is clearly
seen that the ␦Hm values for PBT are lower than the
DSC
values expected on the basis of additivity. One of the
Figures 1 and 2 show DSC heating and cooling scans reasons for this could be the miscibility of PET and
for PET/PBT blends. All of the blends, except the 80:20 PBT, which depends upon the composition. Therefore,
80 ARAVINTHAN AND KALE

Figure 4 Scanning electron microscopy photomicrographs of PET/PBT blends (original magnification ⫻1000): (a) 20:80, (b)
40:60, (c) 50:50, (d) 60:40, (e) 70:30, (f) 80:20, and (g) 80:20 A.

PET-rich blends show different behavior than PBT- phase. DSC studies reported by Avramova7 also
rich blends. show similar results.
Only a single Tg value is observed for all blends. The cooling curves show very interesting behavior.
These values are also listed in Table II. The variation Only one peak is observed for all of the blends. The
of the Tg seems to be linearly proportional to the peak crystallization temperatures are 166°C for PET
change in composition. This also indicates good and 192°C for PBT. The peak crystallization tempera-
miscibility of the two polymers in the amorphous ture for all blends is higher than PET and lower than
BLENDS OF PET AND PBT 81

Figure 4 (Continued from previous page)

PBT. The miscible PBT phase could have a nucleating range of shear rates. It is clear that the viscosity of PET
effect on PET, which can alter these temperatures. Tao is higher than PBT. All blends show Newtonian be-
et al.’s13 results also support the miscibility of PET and havior at a low shear rate and non-Newtonian behav-
PBT and the nucleating effects. ior at a high shear rate. The viscosity of all of the
blends changes systematically with the composition.
This may indicate compatibility between PET and
Rheological properties
PBT. Mishra and Deopura14 studied the rheological
Figure 3 shows the variation of the melt viscosity of behavior of PET and PBT blends. They found that
PET, PBT, and different PET/PBT blends over a wide blends containing up to 4% PBT showed higher vis-
82 ARAVINTHAN AND KALE

cosity than PET. They attributed this affect to in- PET. Thus, the 50:50 blend seems to show unique
creased entanglement. However, when the concentra- properties.
tion of PBT was more than 4%, the viscosity of the
blends was less than PET. They suggested phase seg-
CONCLUSIONS
regation as the possible reason.
The difference in the viscosity of PET and PBT in the The impact strength of all PET/PBT blends is higher
present study at a high shear rate is less than that in than that of PET and PBT. The 50:50 blend showed the
the low shear rate region. Because a lower viscosity highest impact strength value. PET and PBT seemed to
ratio of the blend components provides better disper- be miscible over the entire composition range. The
sion in a given blend, blending of PET and PBT at a miscibility of the amorphous phase of PET and PBT
high shear rate gives better uniformity and properties. resulted in a single Tg. The rheological behavior sug-
The results reported by Jin and Tao8 showed that gested mixing of PET and PBT at a high shear rate.
the viscosity of the blend was less than that of PET and PET-rich blends had a fibrillar nature. Cocontinuous
PBT. Lower viscosity values of blends could be due to behavior was observed for of 40:60, 50:50, and 60:40
slip between the two phases because the intrinsic vis- blend compositions.
cosities of PET and PBT in their study were 0.65 and
1.05 dL/g, respectively. Their processing conditions The first author (G.A.) acknowledges the management of
were also different. Amrita Vishwa Vidyapeetham, Coimbatore, for granting
permission to pursue the M.Tech. program at the University
of Mumbai.
Morphological study
Figure 4 shows scanning electron photomicrographs References
for the PET/PBT blends. The 40:60, 50:50 and 60:40
1. Kroschwitz, J. I. Polymers: An Encyclopedic Sourcebook of En-
PET/PBT blend compositions show a cocontinuous gineering Properties; Wiley: New York, 1987.
morphology [Fig. 4(b– d)]. The 50:50 and 40:60 blends 2. Harper, C. A. Handbook of Plastics and Elastomers; McGraw–
seem to show even a fibrillar nature. The PET-rich Hill: New York, 1975.
80:20 [Fig. 4(f)] and 70:30 [Fig. 4(e)] blends show 3. Mascia, L. Thermoplastics Materials Engineering, 2nd ed.; Ap-
plied Science Publishers Ltd: London, 1989.
slightly higher size of the dispersed phase (PBT) com-
4. Mishra, S. P.; Deopura, B. L. Polym Commun 1985, 26, 5.
pared to the size of the dispersed PET phase in the 5. Mishra, S. P.; Deopura, B. L. J Appl Polym Sci 1987, 33, 759.
PBT-rich 20:80 blend [Fig. 4(a)]. There is also a ten- 6. Stein, R. S.; Khambatta, F. B.; Warner, F. P.; Russel, T.; Escazla,
dency toward a fibrillar nature in PET-rich blends A.; Balizer, E. J Polym Sci Polym Symp 1978, 63, 313.
[Fig. 4(f)]. The morphological behavior of PET-rich 7. Avramova, N. Polymer 1995, 36, 801.
8. Jin, H.-F.; Tao, J. Plast Rubber Compos Proc Appl 1991, 16, 45.
blends reported by Jin and Tao8 is similar to the
9. Porter, R. S.; Wang, L.-H. Polymer 1992, 33, 2019.
present study. The morphology of the 80:20 A blend is 10. Fox, D. W.; Wambach, A. D. (to General Electrical Co.). Ger.
very interesting and quite different than the 80:20 Offen. 2,255,654 (1973).
blend prepared by normal mixing of the two compo- 11. Awazu, N.; Oyama, M. (to Toray Industries Inc.). Jpn. Kokai
nents. The fibrillar nature and some network forma- 73,38,343 (1973).
12. Hrach, J.; Breitenfellner, F. (to Ciba–Geigy A.-G.). Ger. Offen.
tion are clearly visible [Fig. 4(g)]. In the 80:20 A blend,
2,320,117 (1973).
the dispersion of a cocontinuous 50:50 blend into PET 13. Tao, J.; Jin, H.-F.; Sun, T. Plast Rubber Compos Proc Appl 1991,
produced different morphology in the final blend 16, 49.
compared to dispersing 20 parts PBT into 80 parts 14. Mishra, S. P.; Deopura, B. L. Rheol Acta 1984, 23, 189.