The Influence of Thermo-Mechanical History on Structure Development of Elastomeric and Amorphous Glass Thermoplastic Polyurethanes

Transcription

1 The Influence of Thermo-Mechanical History on Structure Development of Elastomeric and Amorphous Glass Thermoplastic Polyurethanes Jorge Silva, 1 Donald Meltzer, 2 Jia Liu, 1 Mark Cox, 2 Jo~ao Maia 1 1 Department of Macromolecular Science and Engineering, Case Western Reserve University, Cleveland, Ohio Lubrizol Advanced Materials, Inc., 9911 Brecksville Road, Cleveland, Ohio In this work, the effect of processing and subsequent thermal treatment on the rheological behavior and microstructure aggregation of two different types of thermoplastic polyurethanes (TPUs) are investigated. Elastomeric TPUs, which are segmented block copolymers composed of alternating soft and hard segments, and amorphous glassy TPU, predominantly composed of hard segments, were subjected to heating/cooling cycles after extrusion. All materials show rheological hystereses after the first thermal cycle; in amorphous glassy TPUs, there is an increase in dynamic moduli but the loss tangent remains unchanged, whereas in the elastomeric TPUs, there are both decreases in dynamic moduli and hystereses in the loss tangent. The changes in molecular weight during the thermal cycle were tracked by gel permeation chromatography and shown not to be responsible for the rheological hystereses. For both elastomeric and amorphous glassy TPUs, fourier transform infrared spectroscopy results indicate that an increase in dynamic moduli, especially G 0, is associated with a higher degree of hydrogen bonding. These results combined with X-ray data indicate that upon extrusion the elastomeric TPUs form a network in which hydrogen bonding plays an important role. This structure can be disrupted upon subsequent heating and cooling. POLYM. ENG. SCI., 54: , VC 2013 Society of Plastics Engineers INTRODUCTION Thermoplastic polyurethanes (TPUs) are a very important and versatile class of polymers. Depending on the details of the chemistry and of the synthesis route, their Correspondence to: Jo~ao Maia; joao.maia@case.edu Jorge Silva is currently at Department of Material Engineering, Federal University of S~ao Carlos, S~ao Carlos, SP , Brazil. Contract grant sponsors: Lubrizol Advanced Materials, Inc. and CLiPS Center for Layered Polymeric Systems, an NSF Science and Technology Center. DOI /pen Published online in Wiley Online Library (wileyonlinelibrary.com). VC 2013 Society of Plastics Engineers final properties can be modulated to cover a wide range of applications, such as medical implants, breathable membranes, cable jacketing, and flexible tubing, to name but a few. The TPUs are multiblock copolymers typically that consist of hard and soft segments. The hard segments (HSs) are composed of diisocyanate and short-chain diols known as chain extenders that form a crystalline phase at room temperature, whereas soft segments (SSs) are, in general, polyethers or polyesters that determine the hydrolytic and oxidative stability of the materials. Hard and soft segments become incompatible at low temperatures, thus resulting in microphase separation in which case the hard domains behave both as crosslinking and reinforcing units, whereas the soft phase is responsible for material flexibility [1]. As such, the physical and mechanical properties of TPUs depend on several factors, including the chemistry and compatibility of soft and hard segments, the ratio of soft and hard blocks, the average block lengths used and their molecular weight distribution. Thermal as well as flow histories can greatly affect the micro-aggregation structure and properties of TPUs [2 6]. Much effort has been spent on investigating thermal transitions and morphological changes in TPUs using different experimental techniques such as differential scanning calorimetry (DSC) [2, 4, 5, 7 9], infrared spectroscopy [2, 10], X-ray scattering [7, 10 14], and nuclear magnetic resonance spectroscopy [2, 5, 13]. Some relatively recent studies [4] have shown that rheology is also a very sensitive tool to investigate the microstructural changes taking place during the phase transition stages. TPUs can exhibit a rich variety of transitional features, ranging from order disorder transition, typical of block copolymers, to microphase separation, and HS solidification [3]. Thermal history greatly affects TPU microstructure formation. For example, hysteresis during heating/cooling processes is frequently observed [3, 5] and annealing greatly affects TPU rheological properties. These POLYMER ENGINEERING AND SCIENCE 2014

2 phenomena seem to be associated with the formation of hydrogen bonding; for example, the time evolution of G 0 during isothermal annealing has been attributed to it [2]. However, the microstructural changes occurring during the annealing or heating/cooling are not as yet wellunderstood. There are also indications that a strain, applied during or just before the structuring process, clearly modifies the kinetics of this structuring and possibly, the resulting morphology [6, 11, 15]. In particular, Mourier et al. [11], studying TPUs with different chemical compositions, reported that a preshear treatment largely diminishes the structuring time. Furthermore, that same preshear treatment becomes efficient above a critical value of shear rate, one that depends on the material nature. In this work, the influence of the thermo-mechanical processing history on morphology and final properties of aromatic TPUs are studied. The materials are extruded into sheet, a process that involves flow and high cooling rates. We focus on the effect of this complex thermomechanical history on the structuring of the material and in particular on the role of HSs in the structural development. This is accomplished by studying two different types of TPUs: amorphous glassy TPUs that are predominantly composed of HSs, and elastomeric TPUs that segmented block copolymers composed of alternating soft and hard segments. Whereas the behavior of elastomeric TPUs is relatively well-documented, studies about amorphous glass TPUs are very rare. One of the few examples is that of Lu et al. [16] who investigated the rheological properties of an aromatic TPU composed exclusively by HSs and observed that the contributions of flow and the degradation reaction of TPU to overall activation energy are additive. EXPERIMENTAL SECTION Materials and Sample Preparation Four commercial materials were mainly used in this work. Two were amorphous glassy aromatic TPUs composed almost exclusively of HSs, Isoplast VR ETPU 301 and Isoplast ETPU These materials differ in the chain extenders from which they are synthesized. The other was elastomeric aromatic TPUs composed of hard and soft ester-based segments. Estane VR TPU and Estane TPU are of similar chemical composition but differ in Durometer, that of former being 85A on the Shore scale and that of the latter being 93A. Estane TPU X1175 is of similar hardness to Estane TPU but is prepared from a different type of polyester polyol. Some of these materials are experimental and all of them are property of Lubrizol Advanced Materials, Inc. For this reason, some details about its synthesis and molecular architecture cannot be provided. For all materials, a film was extruded according to the conditions in Table 1 on an 1 1 = 00 2 single screw extruder. TABLE 1. Extrusion temperature profiles used in this study. Zone 1 Zone 2 Zone 3 Zone 4 Adapter Die Material ( C) ( C) ( C) ( C) ( o C) ( C) Isoplast VR ETPU Isoplast VR ETPU Estane VR TPU X Estane VR TPU Estane VR TPU Rheological Measurements To perform the rheological experiments, 20 mm diameter disks were cut from the extruded film. Prior to the experiments, all samples were vacuum dried for 8 h at 70 C. Rheological measurements were carried out on a rotational rheometer (Thermo MARS III) equipped with parallel plates of 20 mm diameter. Sample thickness ranged between 0.7 and 0.9 mm. During temperature sweep experiments, the normal force was kept constant to compensate for sample density changes. All rheological experiments were conducted under a nitrogen atmosphere to preclude or at least severely delay oxidative degradation of the samples. Samples for gel permeation chromatography (GPC) tests were also collected from the rheometer. The sample collection procedure consisted of stopping the rheometer, quickly removing the sample and quenching it in liquid nitrogen. To evaluate, the thermal stability oscillatory tests were performed over time at 1 rad s 21 at different temperatures. The frequency sweeps were conducted at 175 C with the angular frequency ranged between 0.01 and 100 rad s 21. Temperature sweeps were performed at 1 rad s 21 between 130 and 220 C using different temperature gradients, 3 and 10 C/min. After loading and before experiment, the specimen is hold at 130 C for 10 min. Gel Permeation Chromatography Polymer molecular weight was determined by size exclusion chromatography using a Waters GPC system equipped with an LDC/Milton Roy max N series UV detector. The measurements were taken at 25 C with High Performance Liquid Chromatography-grade tetrahydrofuran (THF) as a mobile phase on two Phenomenex Phenogel columns (100 and 10 nm). Molecular weight was calculated using a calibration based on monodisperse polystyrene standards. Differential Scanning Calorimetry (DSC) The thermal behavior of TPU sheets was determined using DSC (model Q-100, TA Instruments). DSC thermograms were obtained under a nitrogen atmosphere using the same temperatures range and heating/cooling rate (3 C/min) as used in rheology measurements POLYMER ENGINEERING AND SCIENCE 2014 DOI /pen

3 Infrared Spectroscopy In situ fourier transform infrared spectroscopy (FTIR) measurements were performed in a Bomem Michelson MB100 FTIR spectrometer, equipped with a deuterated triglycine sulfate detector and a dry air purge unit. Coaddition of 32 scans was recorded at a resolution of 4 cm 21. Transmission spectra were obtained by casting a thin film on a KBr plate. A hot cell which works as a heating device was adapted to the FTIR machine. The cell assembly was placed in an aluminum block containing two 100-W cartridge heaters designed to maintain a constant temperature through regulation by a proportional-integral-derivative controller and a thermocouple. The cell design allowed for temperature control to within 60.5 C. A small sample of TPU was dissolved in THF first and cast as a thin film on a mm KBr transparent crystal plate. After the solvent was completely evaporated, another mm KBr transparent crystal plate was added to sandwich the sample to regulate the sample thickness and preclude the presence of air during the heating process. The temperature was increased with the rate of 3 C/min from room temperature up to 210 C and FTIR spectra were obtained every 10 C. Small-Angle and Wide-Angle X-ray Scattering Two dimensional (2D) small angle X-ray scattering (SAXS) was performed on a Rigaku SAXS instrument with an 18 kw rotating anode X-ray generator (Micro- Max ) equipped with a Cu tube operated at 45 kv and 0.88 ma. 2D wide-angle X-ray diffraction (WAXD) experiments were performed on an Oxford Xcalibur diffractometer with an ONYX CCD area detector. The X- ray wavelength (k) was Cu KR nm. One dimensional (1D) WAXD profiles were obtained by integrating the corresponding 2D WAXD images. The d-spacing was calibrated using silver behenate with the first-order reflection at a scattering vector q nm 21, where q 5 (4p sin h)/k with h being the half scattering angle. RESULTS AND DISCUSSION FIG. 1. Dynamic moduli of four TPU s (Estane VR TPU X1175, Estane VR TPU 58277, Isoplast VR ETPU 301, Isoplast VR ETPU 2530) measured at a constant frequency (1 rad s 21 ) at 175 C in a nitrogen atmosphere. Shear Rheology and GPC As TPUs can exhibit several phase transitions, each of which can change the rheological properties, a time sweep was performed prior to the frequency sweeps to assess thermal stability of the materials; 175 C was chosen to perform viscoelastic linear measurements as at this temperature, all the materials show a nearly constant behavior over a long enough time period for the frequency sweeps to be completed (Fig. 1). As will be shown below, this temperature is high enough to erase the thermo-mechanical history upon processing. Figure 2 shows the dynamic moduli and the loss tangent of four of the TPUs at 175 C. Contrarily to elastomeric TPUs, in amorphous glassy TPUs, the crossover point occurs at relatively low frequencies and a large rubbery plateau can be observed, mainly in Isoplast ETPU 301. Nevertheless, the loss tangent curves (Fig. 2c) show similar slopes for all materials. As aforementioned, thermal hysteresis is frequently observed in TPUs. In this work, the characteristics of thermal hysteresis in extruded materials were studied. Figure 3 shows the evolution of the storage modulus of elastomeric TPUs measured during a heating and cooling cycle between 130 and 220 C. Both materials exhibit two hystereses in the storage modulus, one at temperatures below 160 C (Estane TPU X1175) or 170 C (Estane TPU 58277), and one at temperatures above approximately 200 C. At low temperatures, the observed hystereses consist in a decrease of dynamic moduli after the completion of the heating and cooling cycles when compared to the initial values. In the loss modulus, which is not show here, Estane TPU X1175 does not exhibit any hysteresis, whereas a small hysteresis at low temperatures is observed for Estane TPU The origin of this behavior as well as the loss tangent will be discussed later. The amorphous glassy TPUs also show the lowtemperature hysteresis in G 0, but not the high-temperature hysteresis (Fig. 4). However, contrarily to elastomeric TPUs, in the amorphous glassy TPUs, the hysteresis consists of an increase in the dynamic moduli after heating/ cooling. The hystereses start at different temperatures in both amorphous glassy TPUs: 220 C for Isoplast ETPU 301 (Fig. 4a) and 200 C for Isoplast ETPU 2530 (Fig. 4b). The loss moduli, which are not shown here, also exhibit hysteresis following the same trend as G 0. That will become clear latter in the discussion of loss tangent. Isoplast ETPU 301, being comprised of a different chain extender than Isoplast ETPU 2530, exhibits a higher Tg than Isoplast ETPU 2530 so it is not unexpected that the low-temperature hystereses would reflect this difference. DOI /pen POLYMER ENGINEERING AND SCIENCE

4 FIG. 3. Storage modulus at 1 rad s 21 measured during temperature sweeps. (a) Estane VR TPU X1175 and (b) Estane VR TPU In all heating and cooling ramps, the temperature gradient was 3 C/min. FIG. 2. Linear viscoelastic material functions for four TPU s (Estane VR TPU X1175, Estane VR TPU 58277, Isoplast VR ETPU 301, Isoplast VR ETPU 2530) at 175 C. (a) Storage modulus, (b) loss modulus, and (c) loss tangent. [Color figure can be viewed in the online issue, which is available at If a second heating/cooling cycle is performed for all materials at the same rate of ramping the temperature of 3 C/min [see Fig. 5, for examples of one elastomeric (Estane TPU 58277) and one amorphous glassy (Isoplast ETPU 301) TPU], G 0 and G 00 (not shown here) superpose onto the curves of the first cooling, that is, the hysteresis disappears after the first heating/cooling cycle, indicating that whatever structure was formed is now stable and reversible. The obvious candidate for these changes in rheological behavior is a variation of molecular weight upon the first heating/cooling cycle. In fact, it has been observed that at high temperatures, the molecular weight of TPUs decreases due to the dissociative reaction of the urethane bonds, whereas it is also well-documented that at lower temperatures, the molecular weight tends to build back up due to formation of new urethane bonds [2, 6, 17 19]. Table 2 presents just such an evolution in molecular weight upon heating and cooling of the TPUs used in this study. Both grades of Estane TPU decrease in the molecular weight at high temperatures and completely recover their molecular weight at low temperatures. Contrary to the elastomeric TPUs, in both of the amorphous glasses, the molecular weight lost during heating is not recovered upon cooling. Moreover, the molecular weight distribution narrows. This indicates that more urethane bonds are cleaved during heating than are reformed upon subsequent cooling. Several degradation mechanism of urethane bonds have been described in literature [17] and the propensity for a particular mechanism depends on the chemical nature of the groups adjacent to the urethane linkage and the environmental conditions. However, the scope of this article is not to discuss the particular degradation mechanism but instead to verify whether the molecular weight variation can explain the rheological behavior. In this respect, our results clearly show that molecular weight variation by itself cannot be responsible for the rheological behavior in temperature sweeps particularly the hystereses at low temperatures, as the decrease in average molecular weight and narrowing of molecular weight distribution observed for amorphous glassy TPUs cannot lead, by themselves, to an increase in dynamic moduli POLYMER ENGINEERING AND SCIENCE 2014 DOI /pen

5 FIG. 4. Storage modulus at 1 rad s 21 measured during temperature sweeps. (a) Isoplast VR ETPU 301 and (b) Isoplast VR ETPU In all heating and cooling ramps, the temperature gradient was 3 C/min. FIG. 5. Storage modulus at 1 rad s 21 measured during two thermal cycles. (a) Estane VR TPU X1175 and (b) Isoplast VR ETPU 301. In all heating and cooling ramps, the temperature gradient was 3 C/min. [Color figure can be viewed in the online issue, which is available at The hystereses at high temperatures during the first heating are observed for both grades of Estane TPUs (see Fig. 6) and are probably due to depolymerization and repolymerization reaction occurring at those temperatures. In Estane TPU X1175, the storage modulus is higher in the initial stages of cooling than in the final stages of heating, the two eventually becoming similar when the temperature once again reaches 200 C. As a consequence of the molecular weight evolution discussed above, at elevated temperatures, some of the urethane bonds cleave, regenerating isocyanate and hydroxyl moieties. During the initial cooling cycle, the isocyanate and hydroxyl moieties (from the chain extender and polyol) react to reform urethane bonds. These reactions can do so in a way that results in a different distribution of hard block lengths than what originally formed when the TPUs were first synthesized. Li et al. [20] and Yamasaki et al. [21] have shown that the distribution of hard block lengths is affected by a variety of factors related to the reaction conditions under which the TPU is synthesized. Thus, it is not unreasonable to suggest that a consequence of the heating cycle, and the associated disruption of hard block structure via the cleavage of urethane bonds followed by the reforming of the urethane bonds under different shear conditions and temperatures under which they were first formed, would result in a different distribution of hard TABLE 2. Molecular weight measured before rheological experiments, after the first heating (220 C) and after a complete thermal cycle (heating and cooling). Extruded sheet No thermal treatment Heated until 220 Cat3 C/min Heated until 220 C at 3 C/min and cooled until 130 Cat3 C/min M w (g/mol) M w /M n M w (g/mol) M w /M n M w (g/mol) M w /M n Isoplast VR ETPU Isoplast VR ETPU Estane VR TPU X Estane VR TPU DOI /pen POLYMER ENGINEERING AND SCIENCE

6 FIG. 6. Loss tangent at 1 rad s 21 measured during temperature sweeps. (a) Estane VR TPU X1175 and (b) Estane VR TPU In all heating and cooling ramps, the temperature gradient was 3 C/min. [Color figure can be viewed in the online issue, which is available at FIG. 7. Loss tangent at 1 rad s 21 measured during temperature sweeps. (a) Isoplast VR TPU 301 and (b) Isoplast VR ETPU In all heating and cooling ramps, the temperature gradient was 3 C/min. [Color figure can be viewed in the online issue, which is available at block lengths. In the second heating/cooling cycle, no hysteresis is observed at high temperatures (Fig. 5a), indicating that the distribution of hard block length is not changing. In the elastomeric TPUs, the hystereses at low temperatures are also observed in the loss tangent (Fig. 6). When remolten in the rheometer, the extruded sheets of Estane TPU initially exhibit a higher storage modulus than loss modulus. However, after heating and cooling cycles, this behavior inverts and the loss tangent increases so as to become greater than one. This suggests that the elastomeric TPUs are changing from a gel-like structure upon processing (including quenching in the rollers) to a liquid-like structure upon subsequent heating and cooling cycles. Gel-like behavior in TPUs has been attributed to the formation and growth of HS ordered/crystalline aggregates, which act as crosslinking nodes [6]. Moreover, in elastomeric TPUs, the cooling curves are smooth at low temperatures, whereas the G 0 curves for the first heating show a region with a fast decrease indicating the presence of a phase transition. Remarkably, in amorphous glassy TPUs, which do not have a significant amount of SSs, there is no hysteresis in tan d, whether at high or low temperatures (Fig. 7). It is thus reasonably to think that the characteristics of the G 0 and G 00 hystereses at low temperatures in the elastomeric TPUs are related to the ratio of hard and soft segments and to how the quenching upon processing affects the phase-separation kinetics. To test this hypothesis, a new sample of film of another elastomeric TPU was prepared, one from Estane TPU As aforementioned, this material is of similar chemical composition to Estane TPU but is of a lower durometer. The experimental results (Fig. 8) show that the hysteresis starts at lower temperature and is less extensive than that of Estane TPU Thus, the extent and the temperature at which the hystereses start clearly seem to depend on the durometer of the TPU and hence on the ratio of hard and soft segments. The actual microstructural phenomena that give rise to this behavior will be discussed below. To confirm that these were not thermally induced kinetic effects, temperature sweeps at 10 C/min were also performed. The results (not shown here) demonstrated that at least in the range of rates at which the temperature was ramped, the phenomena associated with the hystereses at low frequencies are not dependent on the rate at which the temperature was changed. DSC The occurrence of phase transitions that change the microstructure of the material frequently leaves a trace in 1388 POLYMER ENGINEERING AND SCIENCE 2014 DOI /pen

7 FIG. 8. SAOS for Estane VR TPU (a) Storage modulus and (b) loss tangent at 1 rad s 21 measured during temperature sweeps. In all heating and cooling ramps, the temperature gradient was 3 C/min. DSC. Figure 9 shows the DSC thermograms for Estane TPU X1175 and Isoplast ETPU 301 obtained following the same procedure as used in oscillatory temperature sweeps (temperature ranged from 130 to 220 C with 3 C/ min). During the first heating, Estane TPU X1175 shows a transition between 160 and 170 C, whereas in second heating, no thermal transitions are observed. Conversely, no transitions are observed during cooling in the temperature interval at which the rheological experiments were conducted. This is in good agreement with the rheological results. It is thus reasonable to conclude that this phase transition is the reason for the low-temperature hystereses observed in dynamic rheological functions. Endotherm peaks occurring at similar temperatures have been attributed to melting of HS domains formed with hydrogen bonds [5]. However, for Isoplast ETPU 301 (Fig. 9b), which is composed almost exclusively of HSs, no thermal transitions are observed in DSC either in heating or cooling. Thus, the hystereses and the correspondent thermal transition occurring for elastomeric TPUs seem to be driven by the thermodynamic incompatibility between hard and soft segments. Nevertheless, the microstructural processes associated to this thermal transition are not clear yet. To shed light on this matter, IR spectroscopy and X-ray diffraction were also used to study these systems. FIG. 9. DSC thermograms for the two thermal cycles applied to (a) Estane VR TPU X1175 and (b) Isoplast VR ETPU 301 using a heating/cooling rate of 3 C/min. [Color figure can be viewed in the online issue, which is available at IR Spectroscopy To investigate, if there is any variation in the hydrogen bonding associated with the hystereses observed in the thermo-rheological response, IR measurements were performed during heating and cooling. The hydrogen bonds in TPUs occur mainly between NAH and C@O groups. As in the polyester-based TPUs, the carbonyl group is present on both the soft and hard segments, hydrogen bonding can occur between HSs or between hard and soft segments. However, the absorption bands for ester carbonyl group and urethane carbonyl group are normally very close, making them difficult to deconvolute. Figure 10 shows the example of the FTIR spectra of Estane TPU X1175 in the carbonyl (Fig. 10a) and NAH (Fig. 10b) regions during heating from 30 to 210 C. The band assignments for the NAH region and the carbonyl region were summarized by Srichatrapimuk and Cooper [22] and Goddard and Cooper [23]. The carbonyl absorption band is split into two peaks: (i) one at 1735 cm 21 which is attributed to free carbonyl groups and (ii) another at about 1702 cm 21 which is visible mainly at low temperatures and can be attributed to hydrogenbonded carbonyl groups. The NAH absorption band of Estane TPU X1175 is composed of several contributions, of which the most relevant for the current study are at: (i) 3440 cm 21, which is assigned to free NAH groups and (ii) about 3350 cm 21, which is attributed to hydrogen- DOI /pen POLYMER ENGINEERING AND SCIENCE

8 FIG. 10. FTIR spectra in (a) region and (b) NAH region for Estane VR TPU X1175 at various temperatures during heating from 30 to 210 C. [Color figure can be viewed in the online issue, which is available at bonded NAH groups. In Fig. 10, it can be observed that the relative intensity of hydrogen-bonded NAH and carbonyl groups decreases with increasing temperature. Moreover, these peaks are shifted to larger wavenumbers indicating that the strength of hydrogen bonding is weakened with increasing temperature. The spectra obtained during heating and cooling are compared in Fig. 11. At high temperatures, the hysteresis in the spectra is small. However, at 130 C, there is a significant difference between the peaks obtained during heating and cooling, with the results indicating that hydrogen bonding decreases upon completion of the full thermal cycle. Yoon and Han [2] had observed that a lower degree of hydrogen bonding is associated with a decrease in G 0, which is in good agreement with the present results. However, they studied a polyether-based TPU in which there is no C@OHAN hydrogen bonding between hard and soft segments. As stated above, in the case of the amorphous glassy TPUs, there is no significant amount of SS with which the NAH might form hydrogen bonds and yet an increase of dynamic moduli is observed after heating and cooling. This increase would be expected to correspond to a higher degree of hydrogen bonding between C@O and NAH that can form between urethane moieties. FTIR spectra of Isoplast ETPU 301 during heating are presented in Fig. 12. The peaks are roughly in the same position as those of Estane TPU X1175. Likewise, the FIG. 11. FTIR spectra in (a) C@O region and (b) NAH region for Estane VR TPU X1175 at various temperatures during heating and cooling. relative intensity of the peaks associated with hydrogenbonded NAH and carbonyl groups is decreasing with increasing temperature and the peaks are shifted to larger wavenumbers. However, contrary to Estane TPU X1175, in Isoplast ETPU 301 an increase in the intensity of the peaks associated with hydrogen bonding is seen after heating and cooling (Fig. 13). In fact, the hysteresis in the spectra is noticeable even at relatively high temperatures, for example, 190 C. Moreover, there is also a shift of peaks to lower wavelengths indicating that the strength of hydrogen bonding increases after the thermal cycle. This is in good agreement with previous rheological results in that an increase in hydrogen bonding results in the observed increase in dynamic moduli. The results described above indicate that the increase or decrease of G 0 after the first thermal cycle may be attributed, respectively, to the formation or breaking of hydrogen bonds. This data alone, however, do not lend itself to a clear understanding of what is happening in terms of microstructure. To gain insight into this aspect of these polyurethanes, X-ray scattering was used. X-ray Scattering Figure 14 shows the SAXS results for Estane TPU X1175. SAXS of both the extruded sheet and the sample 1390 POLYMER ENGINEERING AND SCIENCE 2014 DOI /pen

9 FIG. 12. FTIR spectra in (a) region and (b) NAH region for Isoplast VR ETPU 301 at various temperatures during heating from 30 to 210 C. [Color figure can be viewed in the online issue, which is available at FIG. 13. FTIR spectra in (a) region and (b) NAH region for Isoplast VR ETPU 301 at various temperatures during heating and cooling. after one thermal cycle show a peak at about 0.40 nm 21, which corresponds to a long spacing of 16 nm. Moreover, the scattering patterns for Estane TPU X1175 (not shown here) are isotropic and hence do not indicate the presence of any orientation effect due to processing. To access smaller scales wide-angle X-ray scattering (WAXS) was also used (Fig. 15). The WAXS pattern is similar for both samples showing only an amorphous halo. The X-ray results suggest that the morphology of Estane TPU X1175 extruded sheet and after heating and cooling is similar. Thus, the most probable cause to the observed hysteresis at low temperature is the difference in molecular conformations. The microstructure generated during processing is more gel-like, with more hydrogen bonding than the morphology obtained after heating and cooling. A better comprehension of these differences can be achieved by testing the amorphous glassy materials. The amorphous glassy TPUs show isotropic SAXS scattering patterns with no peaks, which is expected as the scattering peaks observed in this region result from microphase separation between hard and soft segments, and these TPUs have little if any of the latter. The 2D WAXS patterns for Isoplast ETPU 301 are anisotropic (Fig. 16). The sheet extrusion process leads to molecular orientation (Fig. 16a). The patterns obtained from the heated and cooled sample are less anisotropic indicating that some orientation is lost during the heating/cooling process (Fig. 16b). The 1D WAXS reveals an extra peak around q 5 11 nm 21 after heating and cooling. Combined with the IR results that show an increase in H-bonding, and rheology that shows moduli increase after heating and cooling, this indicates that more structured aggregates of HSs are being formed in the material. Not only were the cooling rates used during these experiments much lower than that used in extrusion, and such that they would facilitate the formation of more ordered hard domains, but as suggested above, the distribution of HS lengths should differ following the first heating/cooling FIG. 14. Small angle X-ray diffraction patterns of Estane VR TPU X1175. The Y-axes have been shifted for the sake of clarity. [Color figure can be viewed in the online issue, which is available at DOI /pen POLYMER ENGINEERING AND SCIENCE

10 FIG. 15. Wide angle X-ray diffraction patterns of Estane VR TPU X1175. [Color figure can be viewed in the online issue, which is available at cycles. Thus, a more ordered hard domain should also be related with a narrower distribution of HSs lengths. As stated above, contrarily to Isoplast ETPU 301, in Estane TPU X1175, no orientation in visible by X-ray diffraction. This is not surprising as the orientation effects should be more pronounced when only HSs are present in the material. Nevertheless, the microstructure of the extruded material is generated by a combination of flow with a fast cooling of the material in the chilled rollers. This will cause the existence of frozen stresses in the extruded sheet that dissipate upon slowly reheating and recooling the TPU. These frozen stresses are not only defined by the extent of microphase separation, for example, being quenched into a nonequilibrium state but are influenced by the distribution of hard block lengths that are known to be influenced by the temperature [5] and shear environment [21] at which the urethane bonds form. This phenomenon is particularly relevant in the mixed phase of elastomeric TPUs where soft and hard segments coexist. The higher level of hydrogen bonding of extruded elastomeric TPUs, as compared with subsequent thermal treated material, should be attributed to the thermalmechanical history during extrusion that involves fast cooling under flow. In fact, as said above, some authors [6, 11] have reported recently that flow strongly affects the structure development of TPUs. An eventual change in the distribution of hard block length could also play a role on the decrease of hydrogen bonding upon thermal treatment. In the glassy amorphous TPUs, the hysteresis results from an increase of dynamic moduli. This driven by the formation of small clusters of HSs and explains why the resulting structure is stable and reversible after a second set of slow heating and cooling cycles. FIG D WAXS diffraction patterns of Isoplast VR ETPU 301. (a) Extruded sheet, (b) after heating and cooling, and (c) 1D WAXS obtained by integration of 2D patterns. [Color figure can be viewed in the online issue, which is available at 1392 POLYMER ENGINEERING AND SCIENCE 2014 DOI /pen

11 CONCLUSIONS In this work, the effect of processing history and subsequent thermal treatment on the rheological behavior of TPUs has been investigated. Two different kinds of TPUs have been used: elastomeric TPUs that are segmented block copolymers composed of alternating soft and hard segments and amorphous glassy TPUs that are predominantly composed of HSs. The results show that the thermo-mechanical history during extrusion induces the formation of a network structure based in the hydrogen bonding in elastomeric TPUs, which is lost during the heating cycle and is not recovered when the temperature is decreased thereafter. This is indicated from the rheological behavior which shows a strong increase in loss tangent and IR spectrum that confirms a decrease in hydrogen bonding. However, SAXS reveals the same level of phase separation before and after heating and cooling cycles. In the case of the elastomeric TPUs, the changes in the microstructure occur at a molecular level not detectable by X-ray and no significant differences were observed in WAXS patterns. A better understanding of the elastomeric TPUs was attained from the study of amorphous glassy TPUs. In the latter case, the dynamic moduli increases after heating and cooling cycles due to the agglomeration of HSs by hydrogen bonding. Thus, the microstructure with high level of hydrogen bonding generated in the mixed phase upon extrusion is responsible for the initially high values of G 0. ACKNOWLEDGMENTS The authors wish to acknowledge David Keifer for performing some of the rheometrical experiments and Tom Braden for extruding the materials into film. The Isoplast VR ETPU and Estane VR TPU are trademarks owned by the Lubrizol Corporation. VC The Lubrizol Corporation 2012, All Rights Reserved REFERENCES 1. G. Oertel and L. Abele, Polyurethane Handbook: Chemistry, Raw Materials, Processing, Application, Properties, Hanser Publishers, New York (1994). 2. P.J. Yoon and C.D. Han, Macromolecules, 33, 2171 (2000). 3. S. Cossar, D. Nichetti, and N. Grizzuti, J. Rheol., 48, 691 (2004). 4. D. Nichetti and N. Grizzuti, Polym. Eng. Sci., 44, 1514 (2004). 5. S. Yamasaki, D. Nishiguchi, K. Kojio, and M. Furukawa, Polymer, 48, 4793 (2007). 6. E. Mourier, R. Fulchiron, and F. Mechin, J. Polym. Sci. Part B Polym. Phys., 48, 190 (2010). 7. A. Saiani, C. Rochas, G. Eeckhaut, W.A. Daunch, J.W. Leenslag, and J.S. Higgins, Macromolecules, 37, 1411 (2004). 8. A. Saiani, W.A. Daunch, H. Verbeke, J.W. Leenslag, and J.S. Higgins, Macromolecules, 34, 9059 (2001). 9. J.T. Koberstein, I. Gancarz, and T.C. Clarke, J. Polym. Sci. Part B.Polym. Phys., 24, 2487 (1986). 10. Y.X. Wang, M. Gupta, and D.A. Schiraldi, J. Polym. Sci. Part B Polym. Phys., 50, 681 (2012). 11. E. Mourier, L. David, P. Alcouffe, C. Rochas, F. Mechin, and R. Fulchiron, J. Polym. Sci. Part B Polym. Phys., 49, 801 (2011). 12. L.T.J. Korley, B.D. Pate, E.L. Thomas, and P.T. Hammond, Polymer, 47, 3073 (2006). 13. S. Velankar and S.L. Cooper, Macromolecules, 31, 9181 (1998). 14. J.T. Koberstein and A.F. Galambos, Macromolecules, 25, 5618 (1992). 15. H.J. Qi and M.C. Boyce, Mech. Mater., 37, 817 (2005). 16. Q.W. Lu, M.E. Hernandez-Hernandez, and C.W. Macosko, Polymer, 44, 3309 (2003). 17. W.P. Yang, C.W. Macosko, and S.T. Wellinghoff, Polymer, 27, 1235 (1986). 18. T. Hentschel and H. Munstedt, Polymer, 42, 3195 (2001). 19. W. Endres, M.D. Lechner, and R. Steinberger, Macromol. Mater. Eng., 288, 525 (2003). 20. C. Li, J.J. Han, Q.S.A. Huang, H.X. Xu, J. Tao, and X.H. Li, Polymer, 53, 1138 (2012). 21. S. Yamasaki, D. Nishiguchi, K. Kojio, and M. Furukawa, J. Polym. Sci. Part B Polym. Phys., 45, 800 (2007). 22. V.W. Srichatrapimuk and S.L. Cooper, J. Macromol. Sci. Phys., B15, 267 (1978). 23. R.J. Goddard and S.L. Cooper, Macromolecules, 28, 1390 (1995). DOI /pen POLYMER ENGINEERING AND SCIENCE