Thermoplastic copolyether ester elastomer toughened polycarbonate blends

Thermoplastic copolyether ester elastomer toughened polycarbonate blends

Polymer Testing 23 (2004) 645–649 Material Properties Thermoplastic copolyether ester elastomer toughened polycarbo...

189KB Sizes 0 Downloads 32 Views

Polymer Testing 23 (2004) 645–649

Material Properties

Thermoplastic copolyether ester elastomer toughened polycarbonate blends 2. Thermal and rheological studies P. Sivaraman, N.R. Manoj, S. Barman, L. Chandrasekhar, V.S. Mishra, A.B. Samui, B.C. Chakraborty  Naval Materials Research Laboratory (D.R.D.O.), Anand Nagar, Ambernath, Maharashtra 421506, India Received 22 December 2003; accepted 28 January 2004

Abstract Blends of polycarbonate (PC) and thermoplastic copolyether ester elastomer (COPE) in different weight ratios have been prepared using a single screw extruder. The infrared spectroscopic studies show no exchange reactions between the phases. However, the DSC scans indicate the alteration of the crystallinity of COPE and a partial miscibility. This is attributed to the migration of PC oligomers, present in PC or generated while mixing, into the COPE phase. Dynamic mechanical analysis also indicates a shift in the glass transition temperatures in the blends. The melt viscosity of the blends is substantially lower than that of pure PC and would help in lowering the processing temperature of PC. # 2004 Elsevier Ltd. All rights reserved. Keywords: Thermoplastic elastomer; Polycarbonate; Toughened blends; Thermal properties; Rheological properties

1. Introduction Polycarbonate (PC), though considered as a tough polymer, may break in a brittle manner, depending on a variety of conditions such as low temperature, notch, thickness of the sample and environmental factors [1–5]. Hence, it is advantageously blended with different polymers to improve the impact strength sensitivity to notch and thickness as well as to develop solvent resistance. However, the blending is frequently impeded by the thermal degradation of the components due to the high processing temperature of PC [1–5]. Blending with a thermoplastic elastomer (TPE) is a recognised route to partly alleviate these problems. Due to its good melt stability and low melt viscosity, TPEs are widely used in polymer blend compositions, parti Corresponding author. Tel.: +91-251-2620608; fax: +91251-2620604. E-mail address: [email protected] (B.C. Chakraborty).

cularly when there is a need to improve the impact strength and to increase the processability of high temperature polymers such as PC [6,7]. Studies on the blends of PC and TPEs such as thermoplastic polyurethane [8,9] and poly (ester–ether) elastomer [10] have been reported. Thermoplastic copolyether ester elastomer (COPE) is an important class of TPEs. They are reported to be segmented copolymers with alternating, random-length sequences of either long chain or short chain oxyalkylene glycols connected by ester linkages. The typical morphology and properties are achieved by the presence of dual phase domains of hard and soft segments in the TPE matrix [6,7]. Short chain cyclic ester units such as tetramethylene terephthalate form hard segments, which are capable of crystallisation with high melting. The soft segments are derived from aliphatic polyether glycols, which form the amorphous blocks. The hard segments provide crosslinking and reinforcement to the soft segments. By varying the composition of soft and hard phases, the properties can be varied over a wide range. In general, thermoplastic TPEs com-

0142-9418/$ - see front matter # 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.polymertesting.2004.01.012


P. Sivaraman et al. / Polymer Testing 23 (2004) 645–649

bine the processability of thermoplastics with the functional performance and properties of conventional thermoset rubbers [6,7]. In an earlier paper, we have reported the mechanical properties and morphology of PC/COPE blends [11]. This work studies the thermal, thermo-mechanical, rheological and infrared spectroscopic characterisation of the PC/COPE blends.

applied on the sample was such as to produce a dynamic (oscillatory) strain amplitude of 32 lm at a fixed frequency of 1 Hz. Melt viscosity of the blends was studied using a Brabender Lab Station capillary die extrusion setup. The length and diameter of the capillary was 15 and 1 mm, v respectively. The temperature of melt was 265  1 C.

2. Experimental

3. Results and discussion

2.1. Materials

It has been observed in the earlier study that the blend containing 20 wt% COPE gives the maximum mechanical properties and impact strength and this has been attributed to the development of characteristic blend morphology [11].

Polycarbonate (Lexan, bisphenol-A type resin) used in this study was procured from GE Plastics, India. Thermoplastic copolyether ester, COPE (Hytrel 4069) was obtained from DuPont, India. PC and COPE were v predried at 120 C for 5 h and 2 h, respectively in an air-circulating oven before the blend preparation. 2.2. Blend preparation A Brabender (Lab Station, Germany) single screw extruder was used for the melt blending of PC and COPE. The detailed procedure of blending is given elsewhere [11]. The PC/COPE blend compositions were 100/0, 90/10, 80/20, 70/30, 60/40 and 0/100 in weight ratios and were coded as PC/COPE0, PC/COPE10, PC/COPE20, PC/COPE30, PC/COPE40 and PC/ COPE100, respectively. The numerical number denotes the weight percentage of COPE in the blend. The extrudate was obtained as 0:5  0:1 mm thin sheets v and were subsequently compression moulded at 250 C in a moulding press at 20 MPa pressure to get sheets of thickness 1  0:2 mm. Test specimens for different characterisation techniques were cut from these sheets. 2.3. Characterisation Infrared (IR) spectra of the blends were recorded on a Perkin–Elmer (model 1600) FTIR spectrophotometer. Thin films of each blend having a thickness of 200 lm was prepared by compression moulding between two v stainless steel plates at 250 C for 2 min. at a pressure of 25 MPa. For the spectra, 16 scans of 4 cm1 resolution were signal averaged in each case. The differential scanning colorimetric (DSC) studies were done using a Modulated DSC (model Q100) of TA Instruments, USA. The samples were scanned in v closed aluminium pans from 30 to 300 C at a heating v rate of 10 C/min with nitrogen purge. The dynamic mechanical analysis (DMA) of the samples was carried out on a Rheometric Scientific Dynamic Mechanical Thermal Analyser model PL Mk v III. The samples were scanned from 100 to +200 C v at a heating rate of 3 C/min. The dynamic stress

3.1. IR studies The IR spectra of the blends and the individual polymers are shown in Fig. 1. COPE is characterised by absorption peaks at 3000–2800 cm1 (aliphatic C–H), 1723 cm1 (carbonyl CAO) and 1080–1125 cm1 (ether C–O–C). PC shows characteristic absorption peaks at 3050, 1010, 829 cm1 (aromatic C–H), 2926 cm1 (hydrocarbonate), 1768 cm1 (carbonyl CAO) and 1600, 1500 cm1 (aromatic ring). Bisphenol-A PC is well known to undergo transesterification reactions with a large number of polymers [12–15]. A transesterification reaction is a necessary requirement for miscibility of polyesters, but requires some degree of compatibility. However, macrophase separation in blends will limit the reaction only to the interfaces leading to either a broadening of Tg or slight shift in Tg [15,16]. The absence of any changes in the absorption peaks indicates that there is no transesterification reaction in the blend. PC– polycaprolactone (PCL) system also does not show any exchange reactions [17], except in the presence of organotin or titanium catalysts [18]. Even in a PC/PET

Fig. 1. Infrared absorbance spectra of the blends (a, pure PC; b, PC/COPE10; c, PC/COPE20; d, PC/COPE30; e, PC/ COPE40; f, pure COPE).

P. Sivaraman et al. / Polymer Testing 23 (2004) 645–649


Table 1 Properties of PC/COPE blends Blend


IR peak (cCAO) (cm1)




Melting v temperature ( C)

Heat of melting/g of COPE in the blend (J/g)

Tan d peak v temperature ( C)

– 1770 1774 1781 1778 1768

1723 1717 1717 1717 1717 –

191.8 190.3 189.1 188.8 186.7 –

14.2 12.3 11.4 10.3 11.7 –

51.1 152.7 154.8 155.9 157.3 161

system, the transesterification reaction has been found to be very slow [19]. The blends, however, display a variation of the peak position corresponding to the carbonyl absorption of PC, as shown in Table 1. It may be seen that the peak position increases gradually from 1768 cm1 in pure PC to 1781 cm1 in PC/COPE 20 and then decreases to 1774 cm1 in PC/COPE 40. The shift in the peak position may be attributed to the presence of some oligomers outside the PC phase, which due to lack of intermolecular interactions, absorb at a higher wave number. The highest shift is seen in the case of PC/ COPE 20, which has shown the highest mechanical properties [11] and hence confirms the concept of maximum penetration into the COPE phase as oligomers. 3.2. DSC studies The DSC thermograms of the pure polymers and the blends are shown in Fig. 2. PC is an amorphous polymer and shows the glass transition temperature v (Tg) at 146 C. The crystalline melting of pure COPE is



observed at a temperature of 192 C. The Tg of COPE is not clearly visible as is the case with many TPEs [6,7]. The suppression of glass transition by the crystalline region is also visible in the dynamic mechanical studies explained later on. The crystalline melting temperatures and the heat of melting for the blends are given in Table 1. The blends show characteristic melting behaviour. The crystalline melting temperatures for the blends shift to lower temperatures as compared to that of pure COPE. The heat of melting per gram of COPE in the blends is also decreased in the blends compared to that of pure COPE. The depression in the melting temperature and the reduction in the degree of crystallinity can be attributed to the diffusion of low molecular weight species to the COPE phase. It has been shown that during blend preparation, low molecular weight oligomers present in PC or generated by the depolymerisation reaction can migrate to the other phase [20–22]. The latter phenomenon has even been shown to result in the lowering of the molecular weight of PC in the blends [23]. The presence of polycarbonates in the COPE phase inhibits the crystallisation process and hence the blends exhibit lesser crystallinity than pure COPE. The shift in the melting temperature is observed to be more in the blends having lower COPE content (i.e. higher PC content). This is due to the possibility of formation and consequent migration of more polycarbonate oligomers from the surplus PC phase. The DMA studies also support this observation of migration of oligomers and the shift in transitions.

3.3. Dynamic mechanical analysis

Fig. 2. DSC thermograms of PC/COPE blends.

The dynamic mechanical properties of the PC/ COPE blends are shown in Figs. 3–5. The dynamic storage modulus (E0 ) as a function of temperature is illustrated in Fig. 3. The magnitude and nature of the change in E0 are dependent on the overall composition of the polymer blends and are determined both by the


P. Sivaraman et al. / Polymer Testing 23 (2004) 645–649

Fig. 3. Dynamic storage modulus (E0 ) vs. temperature of PC/COPE blends.

intermolecular and intramolecular interactions, the latter having greater influence in the different physical states of the polymer [24]. The blends show progressive decrease in E0 as the COPE content increases in the PC matrix. The addition of COPE to the rigid PC matrix increases the flexibility and hence lowers dynamic storage modulus for the blends as compared to pure PC. Pure PC as well as PC/COPE blends show sharp glassto-rubbery transitions whereas pure COPE shows a broad transition. The transition temperature from glassy to rubbery state also decreases gradually with increase in the COPE content. Fig. 4 shows the dynamic loss modulus (E00 ) of the blends and the pure polymers. The intensity of the peak in the loss modulus curve is an indication of the weight fraction of the phases in the polymer blends.

Fig. 5. Dynamic loss factor (tan d) vs. temperature of PC/ COPE blends.

The pure COPE and PC show the respective peaks at v v 68 C and 152 C. With increase in the rubber content in the PC, there is a gradual decrease in the peak heights and also the peaks are observed to broaden and shift to the lower temperature. The glass transition of the amorphous phase of COPE is small as it is suppressed by the crystalline phase. Also in the case of blends, the peaks are almost imperceptible. The loss factor (tan d) as a function of temperature for the PC/ COPE blends is shown in the Fig. 5. The tan d peak values are tabulated in Table 1. Pure PC shows the v tan d peak at 161 C. As the wt% of COPE increases in the PC matrix, the tan d peaks shift to lower temperatures for the blends. In polymer blends, DMA shows a single transition lying in between the individual Tgs if the two components are fully compatible and only one phase exists. On the other hand, if the two polymers are immiscible and exist as two distinct phases, then the blend will show two distinct peaks. If the polymer blend exists in between these two conditions, i.e. they are partially compatible; there will a shift in Tgs towards each other [20]. The shift in Tg in the present case may be attributed to the partial miscibility achieved by the dilution of PC phase by some COPE and, the COPE phase by some PC or PC oligomer. Such cases have been reported in various blends [21–23]. This changes the Tg of the high temperature polymer to the lower side and the low temperature polymer to the higher side.

3.4. Rheological properties

Fig. 4. Dynamic loss modulus (E00 ) vs. temperature of PC/ COPE blends.

The rheological properties of the individual components, while processing, affect the final properties of the product, based on the dispersed droplet sizes [25]. The

P. Sivaraman et al. / Polymer Testing 23 (2004) 645–649


of structure modification in the blends is attributed to the migration of PC oligomers into the COPE phase. The melt viscosity of the blends is substantially lower than that of pure PC and would help in lowering the processing temperature of PC. References

Fig. 6. blends.

Apparent viscosity vs. shear rate of PC/COPE

shear viscosities as a function of shear rate for pure v COPE and PC/COPE blends at 265 C are shown in Fig. 6. At this temperature, it was extremely difficult to carry out viscosity measurement of pure PC since the torque offered by the polymer at this temperature exceeds the capability of the extruder. Pure COPE as well as PC/COPE blends shows linear decrease in the melt viscosity with increase in the shear rate throughout the range of shear rate studied. Pure COPE has much lower melt viscosity than PC/COPE blends. It is evident from the plot that the viscosity of PC/COPE blends decreased almost linearly with increase in the COPE content in the PC. The rheological behaviour suggests that PC/COPE blends can be processed at much lower temperature than pure PC.

4. Conclusion PC was blended with thermoplastic COPE in different weight ratios using a single screw extruder. The blends show partial miscibility through shift in the transitions or melting behaviour, as studied by DSC and DMA. Infrared spectroscopic studies show no exchange reactions between the phases. The mechanism

[1] K.B. Abbas, Polymer 21 (1980) 936. [2] M. Narkis, J.P. Bell, J. Appl. Polym. Sci. 27 (1982) 2809. [3] E.A. Joseph, D.R. Pual, J. Barlow, J. Appl. Polym. Sci. 27 (1982) 2807. [4] L.M. Robeson, S.T. Crisafulli, J. Appl. Polym. Sci. 28 (1983) 2925. [5] M.J. Guest, T.V. Daele, J. Appl. Polym. Sci. 55 (1995) 1417. [6] G. Holden, N.R. Legge, R.P. Quirk, H.E. Schroeder, Thermoplastic Elastomers, Hanser publication, New York, 1996. [7] J.A. Brydson, Rapra Rev. 7 (1995) Report 81. [8] L. Fambri, A. Penati, J. Kolarik, Polymer 38 (1998) 835. [9] K. Palanivelu, P. Sivaraman, S.K. Sharma, Plast. Technol. 5 (2002) 36. [10] K.J. Lee, J.M. Jim, J.G. Kum, J.M. Oh, ANTEC’98, SPE, pp. 2458. [11] P. Sivaraman, N.R. Manoj, S. Barman, L. Chandrasekhar, V.S. Mishra, R. Kushwaha, A.B. Samui, B.C. Chakraborty, Polym. Test. (doi: 10.1016/j.polymertesting.2003.12.001). [12] M.-S. Li, C.-C. Ma, F.C. Chang, Polymer 38 (1997) 855. [13] I. Mondragon, M. Gaztelamendi, J. Nazabal, Polym. Eng. Sci. 26 (1986) 1478. [14] G. Montaudo, C. Puglisi, F. Samperi, J. Polym. Sci., Polym. Che. 31 (1993) 13. [15] H. Ishida, Y.-H. Lee, J. Appl. Polym. Sci. 83 (2002) 1848. [16] O. Becker, G.P. Simon, T. Rieckmann, J.S. Forsythe, R.F. Rosu, S. Volker, J. Appl. Polym. Sci. 83 (2002) 1556. [17] C.A. Cruz, D.R. Paul, J.W. Barlow, J. Appl. Polym. Sci. 23 (1979) 589. [18] M. Shuster, M. Narkis, A. Siegmann, J. Appl. Polym. Sci. 52 (1994) 1383. [19] C. Berty, V. Bonora, F. Pilati, M. Fiorini, Macromol. Chem. 193 (1992) 1679. [20] W.J. MacKnight, F.E. Karasz, J.R. Fried, in: D.R. Paul, S. Newman (Eds.), Polymer Blends, vol. 1, Academic Press, 1978, pp. 185, Chapter 5. [21] M.J. Guest, J.H. Daly, Eur. Polym. J. 25 (1989) 985. [22] J. Mas, A. Vidaurre, J.M. Meseguer, F. Romero, M.M. Pradas, J.L.G. Ribelles, M.L.L. Maspach, O.O. Santana, P. Pages, J. Perez-Folch, J. Appl. Polym. Sci. 83 (2002) 1507. [23] S.B. Kil, O.O. Park, K.H. Yoon, J. Appl. Polym. Sci. 73 (1999) 2123. [24] I. Perepechko, Acoustic Methods of Investigating Polymers, , Mir Publishers, Moscow, 1975. [25] S. Wu, Polym. Eng. Sci. 27 (1985) 335.