Mechanical Properties of Star Block Polymer Thermoplastic

Dec 9, 2016 - Each of the three star polymers exhibited superior recovery (i.e., lower residual strain) and lower hysteresis than the corresponding li...
1 downloads 5 Views 1MB Size
Article pubs.acs.org/Macromolecules

Mechanical Properties of Star Block Polymer Thermoplastic Elastomers with Glassy and Crystalline End Blocks Adam B. Burns and Richard A. Register* Department of Chemical and Biological Engineering, Princeton University, Princeton, New Jersey 08544, United States S Supporting Information *

ABSTRACT: The mechanical properties of thermoplastic elastomers (TPEs) consisting of 6-arm star block polymers with glassy, crystalline, or composite crystalline−glassy physical crosslinking (hard) domains were investigated and compared to the analogous linear triblock or pentablock polymers. The 6-arm stars exhibited qualitatively similar solid-state morphologies and phase behavior to their linear counterparts, as demonstrated by smallangle X-ray scattering and differential scanning calorimetry. Consequently, the architecture had minimal impact on the small-strain behavior in uniaxial extension at room temperature. As the applied strain increased, the star polymers exhibited more pronounced strain hardening than the corresponding linear TPEs, resulting in an increase in the ultimate strength of 20% for the polymers with crystalline end blocks and 30% when the end blocks were glassy. Each of the three star polymers exhibited superior recovery (i.e., lower residual strain) and lower hysteresis than the corresponding linear TPEs when subjected to repeated strain cycles. The enhancement in the recovery was most significant for the polymers with glassy hard domains. The TPEs with crystalline or crystalline−glassy domains recovered more rapidly than the corresponding linear block polymers but showed only modest improvements in the recovery measured after the specimens were allowed to rest for 5 min. These results indicate that the covalent junction at the core of the star strengthens and accelerates the recovery of the network but does not greatly suppress plastic deformation of the crystallites. Overall, this work demonstrates that the mechanical performance of block polymer TPEs can be improved by using a star macromolecular architecture.



INTRODUCTION Thermoplastic elastomers (TPEs) are a versatile class of materials which combine solid-state elastomeric mechanical properties with melt processability. These disparate properties can be realized in a single material by establishing a network of physical cross-links which anchor the rubbery chains at the service temperature but are labile at elevated temperatures.1 This can readily be achieved in A−B−A triblock copolymers in which B is rubbery and the minority A blocks associate to form the physical cross-links (hard domains). Commercial triblock copolymer TPEs employ glassy A blocks. At the appropriate block lengths, chemical incompatibility between the A and B blocks causes them to microphase-separate into discrete, rigid A domains (typically cylinders or spheres for TPEs) embedded in a continuous rubbery matrix of B. When the material is heated above the glass transition temperature (Tg) of the A domains, both phases are liquid-like, but the degree of segregation required to produce high-purity A domains in the solid state typically renders the order−disorder transition temperature inaccessible. The energy barrier for flow in the microphaseseparated melt is large, leading to high viscosities and elasticities compared to typical thermoplastics.2 The processability of linear triblock TPEs can be improved by using semicrystalline end blocks in place of the glassy blocks. Crystallization of the end blocks establishes the requisite network of physical cross-links without the need for strong © XXXX American Chemical Society

interblock incompatibility, thus permitting access to easily processable homogeneous melts above the A block melting point (Tm). In addition, semicrystalline blocks improve solvent resistance. The concomitant gains in processability are offset by the inferior solid-state mechanical performance of polymer crystallites compared to glassy domains.3−5 Crystallization has a strong tendency to produce large, interconnected, plate-like crystallites over a wide composition range. This morphology leads to comparatively high Young’s moduli (EY), and the crystallites are prone to yielding by pullout of individual chains, larger-scale crystal slip, and fragmentation under applied load.6−9 Upon releasing the stress, the initial structure does not recover completely, resulting in substantial permanent set. Structural relaxation is arrested by fusion of the exposed lateral surfaces of adjacent crystal fragments8 and crystallization of oriented segments9 (originating in either the crystalline or amorphous phase). The mechanical properties can be improved without compromising the phase behavior by incorporating both glassy and crystalline blocks to create composite hard domains; one of the simplest block sequences which yields composite domains is crystalline−glassy−rubbery−glassy−crystalline.4,5 IncorporatReceived: October 5, 2016 Revised: November 30, 2016

A

DOI: 10.1021/acs.macromol.6b02175 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Table 1. Molecular and Morphological Characteristics of Block Polymer TPEs polymer

Mw (kg/mol)

Đ

n

(C-EP)2 (C-EP)6 (E-EP)2 (E-EP)6 (E-C-EP)2 (E-C-EP)6

144 427 137 478 158 500

1.08 1.21 1.07 1.15 1.07 1.13

1.95 5.76 1.95 6.13 1.89a 5.86

wE

0.20 0.19 0.09 0.09

wC

d (nm)

peak Tm (°C)

wc,Eb

0.20 0.19

37 37 49 52 44 49

102 98 93 87

0.48 0.50 0.36 0.44

0.10 0.10

a Mw,arm determined on a partially coupled sample. bWeight fraction crystallinity of the E blocks wc,E = ΔHm/(wE × ΔHm,100), where ΔHm,100 = 277 J/ g is the melting enthalpy of 100% crystalline E.35

understand the influence of the branch point on the mechanical response. To this end, we focus on precisely defined 6-arm star block polymers prepared by anionic polymerization, chlorosilane coupling, and catalytic hydrogenation.32 A star functionality of six was chosen based on previous work17,18,21,22 demonstrating that the ultimate stress reaches a plateau in the range of n = 5−10. The core of the star is expected to function as a covalent cross-link which can distribute stress more uniformly and provide additional memory to the network without significantly altering the morphology or melt processability.17−20,28 We report results from a series of three star block polymers having rubbery inner blocks and glassy, crystalline, or composite crystalline−glassy hard domains. The tensile properties and strain recovery behavior of the star block polymers are thoroughly characterized and compared with those of well-matched linear block copolymers.

ing the glassy block in this architecture has been shown to improve (reduce) EY and (increase) σu. Whether the strain recovery is improved by the presence of the glassy block depends strongly on the identity of the crystalline blocks. When the crystalline blocks tended to form large, highly interconnected crystallites, incorporating the adjacent glassy blocks markedly improved the recovery.4 However, when the size and connectivity of the crystallites were inherently low and the behavior was already qualitatively elastomeric, the glassy blocks had essentially no effect on the recovery.5 Plastic deformation of the hard domains,10−14 which leads to yielding and comparatively high residual strain (permanent set) and dissipative loss (hysteresis), remains a significant drawback of TPEs compared to chemically cross-linked elastomers,15 regardless of the identity of the hard blocks. One way to improve the mechanical performance of TPEs irrespective of the type of hard domains might be to use well-defined branched macromolecular architectures. The simplest and most common branched architecture is a star block copolymer comprising A− B arms with the B ends connected to a central core (denoted (A−B)n, where n is the number of arms). Star block copolymers with glassy end blocks, often referred to as “radial” block copolymers in the context of TPEs, have been prepared by a variety of methods.16−28 Commercially, star block copolymers are preferred in some applications where the block copolymer is blended with other components such as in thermoplastic toughening, pressure-sensitive adhesives, and asphalt modification.29 In many instances neat star block copolymers with glassy end blocks have been found to exhibit higher ultimate strength (σu) than their linear analogues,17,18,20−28 which commonly comes at the expense of reduced breaking strains (εb).17,18,22,27,28 In instances where direct comparisons between linear and star block copolymer have been made, reports of both enhanced17 and diminished18 strain recovery (with respect to the corresponding linear polymers) can be found in the literature. Semicrystalline TPEs with well-defined branched architectures have not been studied in detail, but some evidence of improved recovery can nonetheless be found.30,31 Improvements in the mechanical properties in branched TPEs have been attributed to increased connectivity between the rubbery phase and the hard domains, which is better able to distribute stresses through the network.17,18 Moreover, the processability was found to depend primarily on the composition and molecular weight of the arms and only weakly on n for n ≥ 3.17−20 In some cases, the contribution of the covalent cross-links is obscured by changes in the morphology arising from the architectural change,23 broad molecular weight (arm number) distributions,17,24,28 and non-negligible filler effects from the core of the star.17,25,28 Here we attempt to avoid contributions from morphological changes and linking chemistry to better



EXPERIMENTAL SECTION

Polymers were synthesized by sequential anionic polymerization and coupled with either a difunctional (dimethyldichlorosilane) or a hexafunctional (2,2,4,4,6,6-hexachloro-2,4,6-trisilaheptane) chlorosilane to afford precisely defined, symmetric linear or 6-arm star block polymers, respectively. The coupled polymers were subsequently hydrogenated over Pd (5 wt %) supported on CaCO3. The detailed synthetic procedures have been described elsewhere.5,32 The semicrystalline block is hydrogenated high 1,4-polybutadiene (linear lowdensity polyethylene, E), hydrogenated polystyrene (polyvinylcyclohexane, C) serves as the glassy component, and the rubbery block is hydrogenated high 1,4-polyisoprene (poly(ethylene-altpropylene), EP). Block polymers with three types of hard domains were studied: glassy (C-EP)n, crystalline (E-EP)n, and composite crystalline−glassy (E-C-EP)n. The linear TPEs (n = 2), which serve as a point of comparison, have been characterized in detail elsewhere.5 Absolute weight-average molecular weights (Mw) of the polymers prior to hydrogenation were determined by gel permeation chromatography (GPC) in THF using differential refractive index (Wyatt Optilab TrEX) and multiangle light scattering detection (Wyatt miniDAWN TREOS). GPC traces of the unhydrogenated star polymers and hydrogenated (C-EP)6 are presented in the Supporting Information. The functionality of the coupled polymers was calculated by n = Mw,coupled/Mw,arm, where Mw,arm and Mw,coupled correspond to the polymers before and after chlorosilane coupling (including any uncoupled material), respectively. Compositions were determined using 1H nuclear magnetic resonance spectroscopy on the unsaturated precursors; block weight fractions (w, adjusted for the addition of hydrogen) are reported in Table 1 along with other relevant molecular characteristics. Small-angle X-ray scattering (SAXS) was used to characterize the morphology and phase behavior of each polymer. The SAXS system comprised a PANalytical PW3830 generator with a long-fine-focus tube producing Cu Kα X-rays (λ = 1.5418 Å), a slit-collimated Anton Paar compact Kratky camera equipped with a hot stage, and an MBraun OED-50 M position-sensitive detector. The raw data were B

DOI: 10.1021/acs.macromol.6b02175 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Table 2. Mechanical Properties of Block Copolymer TPEs polymer (C-EP)2 (C-EP)6 (E-EP)2 (E-EP)6 (E-C-EP)2 (E-C-EP)6

EYa (MPa) 7.1 7.3 10.2 10.5 6.7 7.1

SDb (%) 34 5 2 5 4 6

EY/E0Nd 2.1 2.1 2.9 3.1 1.9 2.1

σu (MPa) c

SDb (%)

a

11.0 /9.6 16.5/12.8 6.4/6.4 8.2/7.7 8.4/6.9 8.9/7.3

21 36 1 8 20 19

εb (%) c

850 /800 700/660 570/560 580/540 740/660 620/540

a

SDb (%)

Tu (MJ/m3)

SDb (%)

7 9 1 6 15 13

c

22 27 1 14 31 27

a

34.1 /29 34.4/28 19.0/19 23.7/21 29.1/22 26.3/20

Average of at least three specimens. bOne standard deviation of at least three specimens. cMaximum measured value; for (C-EP)2 the maximum σu, εb, and Tu were obtained from three different specimens, while for all other polymers the maximum values were measured in a single specimen. d Young’s modulus of TPE divided by tensile plateau modulus of EP (3.45 MPa).37 a

bars in Figures 2, 5, 7, and 8 represent the average and ±1 standard deviation determined from three test specimens.

corrected for detector sensitivity, empty beam scattering, and sample thickness and transmittance, converted to absolute intensity using a polyethylene standard, and desmeared for slit length.33 The patterns of E-containing block polymers are presented as plots of q2I versus q, where q = (4π/λ) sin θ is the magnitude of the momentum transfer vector; the scattering angle (2θ) was calibrated with silver behenate. This construction approximately corrects for the form factor of lamellae.34 The data for (C-EP)n are presented as qI versus q to approximately correct for the form factor of cylinders.34 The domain periodicity (d) was calculated for each polymer from the location of the primary scattering maximum (q*) determined from a plot of qI versus q for (C-EP)n or q2I versus q for (E-EP)n and (E-C-EP)n, as d = 2π/q*. The semicrystalline polymers were also characterized by differential scanning calorimetry (DSC) using a PerkinElmer DSC 7 equipped with a Type II intracooler calibrated with indium and mercury. A scan rate of 10 °C/min was used throughout. The melting enthalpy (ΔHm) was calculated by integrating the melting endotherms using a baseline formed by extrapolating the melt data to lower temperature until it intersected the solid-state heat capacity, near 25 °C. Samples were cut from the same molded sheets used for mechanical testing. The derived morphological characteristics are listed in Table 1. The mechanical properties were characterized in uniaxial extension at room temperature using an Instron 5865. Compression-molded sheets were prepared using a PHI hydraulic press. Samples were loaded into the press between Mylar sheets sandwiched between aluminum plates. Each sample was heated on the press for 5−10 min and then pressed for 5−10 min. The sample was removed from the press and allowed to cool to room temperature on the benchtop between the aluminum plates (30−45 min). The materials with crystalline end blocks (which exhibit homogeneous melts) were pressed at 140−160 °C and ≤2 MPa. The materials with glassy end blocks (including Kraton D1111) were pressed at 170 °C and 3−6 MPa. The resulting sheets were approximately 300 μm thick. Test specimens were stamped from the compression-molded sheets using an ASTM D1708 dogbone die. Two types of test were conducted: (1) strain to break and (2) cyclic strain recovery. A constant cross-head speed of 2 in./min (corresponding to an initial strain rate of 0.038 s−1) was used throughout. The former test was used to measure EY and the ultimate properties, σu and εb. Additionally, the ultimate toughness (Tu) was calculated by integrating the area under the stress−strain curves, where the strain is in units of mm/mm. The values reported in Table 2 were averaged over a minimum of three specimens. The following cyclic strain test was used to measure recovery: a specimen was strained to a prescribed strain (εa), the cross-head was then returned to its initial position at the same rate, and the specimen was allowed to rest for 5 min before the process was repeated, increasing εa. Strains of 25, 50, 100, 150, 200, 300, 400, and 500% were applied to each specimen; Figure S5 shows an example of the raw data. The residual strain (set, εs) was measured at two time points for each cycle. For a given cycle, the “initial” set corresponds to the strain at which the stress first dropped to zero while the cross-head was returning to its initial position. The “final” set was defined as the strain at which the stress first became nonzero on the subsequent advancing cycle. The percent recovery was calculated by R = (1 − εs/εa) × 100%; initial and final recovery are denoted Ri and Rf, respectively. Data points and error



RESULTS AND DISCUSSION The SAXS patterns of (C-EP)2 and (C-EP)6 at room temperature and 200 °C are shown in Figure 1. The domain

Figure 1. SAXS patterns of (C-EP)2 and (C-EP)6 at room temperature (a) and at 200 °C (b). The (C-EP)6 patterns have been shifted up by a factor of 3 in intensity for clarity.

periodicities of (C-EP)2 and (C-EP)6 were found to be equal after heating to 200 °C and cooling to room temperature (Table 1 and Figure 1). Based on self-consistent field theory predictions,36 the equilibrium morphology of (C-EP)2 and (CEP)6 was expected to be hexagonally packed cylinders of C. The lack of well-developed higher order peaks in the SAXS patterns precluded a definitive morphological assignment; however, the kink in the scattering patterns near q = 0.4 nm−1 agrees well with the form factor minimum for cylinders5,34 (see Supporting Information). Thus, the morphology appears to be cylindrical C domains with poor long-range order. Although the morphology and average periodicity between C domains was equivalent in the linear and star polymers, the morphology of (C-EP)6 was somewhat less regular, as evidenced by the broader primary SAXS peak. Both C

DOI: 10.1021/acs.macromol.6b02175 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

the intrinsic ultimate properties are difficult to measure experimentally due to premature failure induced by impurities. Interestingly, the energies (per unit volume of deformed material) required to break the specimen (i.e., the ultimate toughness) of the linear and star polymers were equal (Table 2). The initial and final cyclic recovery behavior of (C-EP)2 and (C-EP)6 are shown in Figures 2b and 2c, respectively. The star architecture clearly provided enhanced recovery across the range of εa investigated, both in Ri and Rf. In addition to a more even distribution of stresses, the permanent cross-link at the core of the star evidently provides additional memory to the network. Unlike the stress−strain behavior, the improvement in the strain recovery of (C-EP)6, relative to (C-EP)2, diminished as the strain increased. Unfortunately, the recovery of (C-EP)6 could not be accurately determined at εa = 500% because the specimens slipped in the grips during testing. At these large strains the glassy domains have undergone substantial restructuring,11,13 which dominates the residual strain and apparently reduces the ability of the permanent cross-link to restore the initial state. As a point of comparison, the recovery of (C-EP)2 was on par with a commercial styrenic triblock copolymer TPE (see Figure S6) with comparable composition (Kraton D1111, a polystyrene−polyisoprene−polystyrene triblock copolymer with 20 vol % polystyrene, compared to 18% C in (C-EP)2). The SAXS patterns of (E-EP)2 and (E-EP)6 showed a weak primary peak at low q followed by a broad region of intensity extending to higher q arising from the intercrystallite spacing (Figure 3). The morphology of (E-EP)2 and (E-EP)6 consists of clusters of E crystallites separated by regions of amorphous EP in an approximately lamellar arrangement with characteristic spacing d.39,40 The characteristic periodicity of (E-EP)6 was 6% larger than that of (E-EP)2, which is attributed to the slightly (16%) higher arm molecular weight in (E-EP)6. Compared to

polymers remained microphase separated in the melt to at least 200 °C (d = 38 nm). Figure 2a shows stress−strain curves for the linear and 6-arm star TPEs with glassy end blocks in uniaxial extension. The

Figure 2. Mechanical properties of (C-EP)6 and the corresponding linear triblock (C-EP)2: stress−strain curves (a), initial recovery (b), and final recovery 5 min after the stress has been removed (c).

average Young’s modulus (EY), ultimate strength (σu), breaking strain (εb), and ultimate toughness (Tu) measured therefrom are listed in Table 2. At small strains the linear and star polymers behaved similarly since the response is dominated by trapped entanglements (the entanglement molecular weight for EP37 is ≈1.5 kg/mol) and the volume fraction and morphology of the C domains, which were minimally impacted by the change in architecture. The variability in the modulus of (CEP)2 was attributed to differences in the alignment of the cylindrical domains with respect to the stretching direction.11,13,38 Above about 50% strain, (C-EP)6 exhibited more pronounced strain hardening than (C-EP)2, and σu was increased by a factor of 1.3, while εb was 140% lower on average. These results are consistent with previous reports on star block copolymer TPEs with glassy end blocks.17,18,22,27,28 Comparing the best-performing specimens of each material suggests that the true enhancement might be even greater, but

Figure 3. SAXS patterns of (E-EP)2 and (E-EP)6 at room temperature (a) and at 110 °C (b). The (E-EP)6 pattern has been shifted up by a factor of 2 in intensity in (a) for clarity. D

DOI: 10.1021/acs.macromol.6b02175 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules (E-EP)2, the SAXS pattern of (E-EP)6 exhibited a weaker low-q feature, but higher intensity in the intermediate q-range (q = 0.2−0.8 nm−1), indicative of a broader distribution of the crystal−amorphous spacing. The peak melting point of (EEP)6, measured by DSC (Table 1 and Figure 4), was

Figure 4. DSC traces of the semicrystalline linear and star TPEs on the first heating cycle (a) and subsequent cooling (b).

suppressed by 4 °C compared to (E-EP)2indicating that the crystals are somewhat thinner in the star polymerdue presumably to retarded mobility of the star polymer during crystallization. However, the degree of crystallinity of the E blocks (wc,E) in the star and linear polymers is comparable (Table 1). Throughout this work the upper bounds on the crystal thickness and crystallinity of the E blocks are set by the ethyl branches arising from 1,2-units in the precursor polybutadiene blocks; the observed difference in the peak melting points does not stem from differences in the branch content of (E-EP)2 and (E-EP)6 (20.6 and 19.8 branches per 1000 backbone carbons, respectively). Figure 5 shows the stress−strain behavior for TPEs with crystalline end blocks: (E-EP)2 and (E-EP)6. The Young’s moduli of the star and linear polymers were equal within experimental uncertainty (Table 2). In spite of the reduced crystal thickness, which is expected to make the crystals more susceptible to plastic deformation, (E-EP)6 was actually stiffer than (E-EP)2 above ≈50% strain, culminating in an average increase in σu by a factor of 1.2, with no change in εb. The bestperforming sample of (E-EP)6 also showed a factor of 1.2 increase in toughness. Additionally, the 6-arm star outperformed the linear polymer in initial recovery (Figure 5b). Similar to the polymers with glassy end blocks, the difference between (E-EP)2 and (E-EP)6 diminished as εa increased until the recovery was essentially equal for εa ≥ 400%. The recovery benefits of the star architecture also diminished after the samples were allowed to rest for 5 min (Figure 5c). Both the linear and star polymers ultimately recovered to nearly the same degree; however, (E-EP)6 recovered more quickly, with

Figure 5. Mechanical properties of (E-EP)6 and the corresponding linear triblock (E-EP)2: stress−strain curves (a), initial recovery (b), and final recovery 5 min after removal of stress (c).

only ca. 2% of additional recovery during the 5 min rest, while (E-EP)2 recovered nearly twice as much over the same period. These results imply that the covalent cross-link at the center of the star facilitates rapid recovery of the (visco)elastic deformation but does not dramatically suppress the plastic deformation mechanisms which contribute to permanent set. It is noteworthy that (E-EP)6 outperformed (E-EP)2, if only modestly, in spite of the reduced crystal thickness; even better performance might be achieved by tuning the crystallization conditions to obtain comparable crystal thicknesses. Comparing the recovery of (E-EP)2 and (E-EP)6 to those of (C-EP)2 and (C-EP)6 (cf. Figures 2b and 2c), the latter of which were significantly higher, illustrates the adverse effects of plastic deformation in the crystalline domains. As mentioned above, linear polymers with the block sequence crystalline−glassy−rubbery−glassy−crystalline have been demonstrated to mitigate some of the drawbacks of using crystalline hard blocks alone, while maintaining access to homogeneous melts.4,5 Here, this concept is extended to star block polymers with the same motif, namely (crystalline− E

DOI: 10.1021/acs.macromol.6b02175 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules glassy−rubbery)n. Theory41,42 predicts that the phase envelope (order−disorder transition, ODT) asymptotically shifts to lower arm molecular weight as the number of arms increases (i.e., the ordered state is stabilized by the star architecture) but that the dependence of the ODT on n is weak, a point which is corroborated by experiments.43,44 Thus, a single, judicious choice of block lengths should yield single-phase melts in both the linear and 6-arm star block polymers. (E-C-EP)2 and (E-CEP)6 provide a rather stringent test of this hypothesis. SAXS patterns of (E-C-EP)2 exhibited correlation hole scattering just above Tm (Figure 6b), indicating that the polymer was

The mechanical properties of (E-C-EP)6 are shown in Figure 7. As with (C-EP)6 and (E-EP)6, the small-strain behavior of

Figure 6. SAXS patterns of (E-C-EP)2 and (E-C-EP)6 at room temperature (a) and at 110 °C (b). The (E-C-EP)6 pattern has been shifted up by a factor of 2 in intensity in both panels for clarity. Figure 7. Mechanical properties of (E-C-EP)6 and the corresponding linear pentablock (E-C-EP)2: stress−strain curves (a), initial recovery (b), and final recovery 5 min after removal of stress (c).

disordered but close to the ODT. Correlation hole scattering was also observed in the melt-state SAXS patterns of (E-CEP)6, indicating that the phase envelope had not shifted enough to qualitatively change the phase behavior in going from n = 2 to n = 6 (i.e., the order−disorder transition temperature still lies close to but below Tm). The solid-state SAXS pattern of (E-CEP)6 also resembled that of (E-C-EP)2 (Figure 6a), though the characteristic spacing between composite crystalline−glassy domains was 10% larger in the former. DSC revealed that the Tm of (E-C-EP)6 was 6 °C lower than in (E-C-EP)2, but the apparent crystallinity of the E blocks was 20% higher in the star. A portion of this endotherm can be attributed to the convolution of devitrification and dissolution of C block aggregates with melting of the crystallites. Again, the differences in melting point (crystal thickness) and crystallinity cannot be explained by differences in ethyl branch content, which are 20.1 per 1000 backbone carbons in (E-C-EP)2 and 20.3 in (E-CEP)6. The increases in d and the apparent wc,E observed in the star polymer may indicate that the C blocks are better segregated from E and EP in (E-C-EP)6 than in (E-C-EP)2 following E block crystallization, resulting in purer, higher-Tg C domains. This is consistent with a shift in the ODT to lower segregation strength (higher temperature) with increasing n.

(E-C-EP)6 closely mirrored that of its linear analogue. At moderate strains (E-C-EP)6 was stiffer than (E-C-EP)2, again consistent with the behavior of the TPEs with purely glassy or purely crystalline end blocks. The enhanced stiffness of (E-CEP)6 did not result in a significant improvement in σu, and the differences in εb and Tu were well within the uncertainty of the measurement. Again, the true ultimate properties can be difficult to measure experimentally, so it is unclear whether this behavior is intrinsic. (E-C-EP)6 showed better initial recovery than (E-C-EP)2, as was observed for (E-EP)6. Interestingly, Rf was also higher for (E-C-EP)6 compared to (E-C-EP)2, particularly at the lowest εa, despite the finding that neither (E-C-EP)2 nor (E-EP)6 provided significantly better performance than (E-EP)2. The apparent synergistic effects of the glassy blocks and the star architecture may be explained by the formation of more robust C domains in (E-C-EP)6 inferred from the increases in d and the apparent wc,E. In addition to the recovery, another quantitative metric for elasticity is fractional hysteresis (H), which measures the amount of energy dissipated during deformation as heat.15,45 It F

DOI: 10.1021/acs.macromol.6b02175 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

(EY = 7.3−8.2 MPa for the three specimens tested), and there was correspondingly less variability in H. If all of the samples of (C-EP)2 were considered, the star architecture reduced H by 30%. (C-EP)6 had slightly more hysteresis than the bestperforming specimen of (C-EP)2 at small strains (εa ≤ 50%) but still performed better by a factor of 1.2 for εa ≥ 100%. Turning to the semicrystalline polymers, the hysteresis of (EEP)2 agreed well with prior measurements on an E-EP-E triblock with 18 wt % E.46 The hysteresis of (E-EP)6 was reduced by a factor of 1.5 with respect to that of (E-EP)2 for εa ≤ 200%. Similar behavior was observed for (E-C-EP)6, which showed a factor of 1.3 reduction in H compared to its linear counterpart (E-C-EP)2 for εa ≤ 200%. In the range εa = 300− 400% H for the semicrystalline stars and the corresponding linear polymers were comparable, and at εa = 500% the linear polymers showed less hysteresis. The initial recovery showed a similar trend; namely, the improvements exhibited by the 6-arm star diminished as εa increased. However, Ri for the stars never fell below that of the corresponding linear polymer, and in the case of the composite hard domains (E-C-EP)6 exhibited better Ri across the entire range of εa. For comparison, H ≈ 0.2 at εa = 100% for vulcanized natural rubber (NR) and styrene− butadiene rubber (SBR) with 20 wt % filler.45 It bears mentioning that the hysteresis of vulcanized rubber depends strongly on the filler content (H ≈ 0.05 for unfilled NR and 0.1 for unfilled SBR), cross-linking density, and testing rate. Nevertheless, all three star polymers studied here provided hysteresis performance which is competitive with that of filled vulcanized rubbers. Furthermore, the improvements in H associated with the star architecture are comparable to improving H in vulcanized rubber by reducing the filler fraction. However, changing the filler fraction changes EY,47 whereas the star TPEs studied here did not. It is clear that the star architecture is an effective way to improve the mechanical performance of TPEs, but the property which was most improved and the degree of improvement was found to vary depending on the identity of the end block. During the initial deformation (less than 50% strain) entanglements dominate, and the contribution of the core of the star is minimal; hence EY changes very little. At moderate and large strains, the role of the covalent junction at the core of the star is to more effectively distribute stresses to the hard domains, 17,18 which makes the stars stiffer than the corresponding linear polymers at any given strain (above ≈50% strain). In this regime, chain pullout and deformation of the hard domains are expected to limit the recovery, contribute to the hysteresis, and ultimately lead to failure for all of the polymers studied here.14,48−50 The role of the star architecture is to maintain elastically effective connections between the rubbery phase and the hard domains in spite of these processes. If a hard block becomes detached from its corresponding domain (by pullout, scission, or domain rupture), the triblock copolymer can relax completely, releasing trapped entanglements and contributing to the hysteresis and permanent set (Figure 9a). In the case of the star block copolymer only the failed arm can relax (releasing only half as many entanglements as in the triblock if the arm molecular weight is equivalent) while the intact arms remain anchored to the network (Figure 9b), leading to less hysteresis and better recovery. The same argument holds regardless of the type of hard domain (glassy, crystalline, and composite). The purely crystalline end blocks are the most prone to plastic deformation and thus show improvements in the rate of recovery (Ri) and

is desirable to minimize hysteresis in elastomers, particularly in applications where the material is subjected to cyclic deformation. Fractional hysteresis was calculated from the same cyclic strain data used to compute the recovery by taking the difference in the areas under the loading and unloading curves normalized by the area under the loading curve. Figure 8

Figure 8. Hysteresis for (C-EP)n (a), (E-EP)n (b), and (E-C-EP)n (c). Panel (a) shows H for the best-performing sample of (C-EP)2 as well as H averaged over all three specimens tested.

shows H as a function of εa for each of the six TPEs studied. The hysteresis behavior of (C-EP)2 was complicated by the influence of the morphology. As mentioned above, the mechanical properties of (C-EP)2 are sensitive to the degree of alignment between the microdomains and the stretching direction. The initial modulus of the three specimens used for cyclic testing ranged from 6.6 to 28.7 MPa, and the concomitant changes in the shape of the stress−strain curve11,13,38 resulted in large variability in H (specimens with higher EY exhibited more hysteresis). Nevertheless, the hysteresis of (C-EP)2 was remarkably similar to that of Kraton D1111 (Figure S6). Interestingly, the shape of the cyclic stress−strain curves of (C-EP)2 did not greatly impact the recovery, as evidenced by the comparatively small error bars in Figures 2b and 2c. (C-EP)6 showed no evidence of alignment G

DOI: 10.1021/acs.macromol.6b02175 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

somewhat worse than for the triblocks.52,53 Graft copolymers with EP backbones (synthesized by copolymerizing ethylene and propylene) and crystalline tactic PP grafts provided similar improvements in Rf compared to the conventional triblock architecture as the 6-arm stars studied here. A graft copolymer with syndiotactic PP hard blocks showed an increase in Rf by almost 20% (at εa = 500%) over an sPP-EP-sPP triblock copolymer, which suffered from considerable permanent set. However, the best-performing graft copolymers had Rf ≈ 90%, lower than (E-EP)2 and (E-EP)6.54 Determining the contribution of the graft architecture is also complicated by the fact that the hard block fraction in the grafts was 1200% but had weaker strain hardening and lower σu, in contrast to the star block polymers studied here and by others.17,18,20−28 A polymer with similar hard block composition (17 vol % S) to (C-EP)2 and (C-EP)6 and 3.7 grafts per chain (with one S chain per graft point) exhibited Rf ≈ 93%, less than even (E-EP)2. A graft polymer with 22 vol % S and 5.2 grafts per chain (with four S chains per graft point) had Rf ≈ 95% compared to Rf > 97% for (C-EP)6. This graft copolymer also showed considerable hysteresis, although H was not quantified. While studies of semicrystalline TPEs are comparatively rare, some architectures more complex than triblocks have been reported. Kong et al. demonstrated that a 4-arm star30 with 22 wt % semicrystalline poly(L-lactide) end blocks recovered >10% more than its linear analogue31 (εa = 150%), but Rf was still less than 85%. Coates and co-workers have reported a series of linear multiblock copolymers having alternating crystalline and amorphous blocks.52,53 They found that substantial improvements in σu can result from having central crystalline blocks flanked by rubbery blocks (e.g., iPP-rPP-iPP-rPP-iPP versus iPP-rPP-iPP,52 where rPP and iPP are rubbery regioirregular and crystalline isotatic polypropylene, respectively); however, the recovery of the multiblocks was comparable if not



CONCLUSIONS



ASSOCIATED CONTENT

In conclusion, using a star macromolecular architecture improved the ultimate strength, strain recovery, and hysteresis of block polymer thermoplastic elastomers with glassy, crystalline, and composite crystalline−glassy physical crosslinks. The star architecture did not qualitatively alter the network morphology or phase behavior in any of the three cases studied. As a result, stress−strain curves for the 6-arm star polymers were almost indistinguishable from those of their linear analogues at small strains. At moderate and high strains the core of the star acts as a permanent cross-link, which distributes stresses more evenly and provides additional memory to the network, resulting in enhanced stiffness and recovery. The improvements were most pronounced when the hard domains are entirely amorphous (i.e., glassy). Interestingly, the star TPE with purely crystalline hard domains exhibited higher ultimate stresses and recovered at least as well as its linear analogue despite having somewhat thinner crystals. The behavior of the star TPE with composite hard domains was similar although σu was unchanged. The covalent cross-link accelerated recovery and reduced the hysteresis during cyclic strain in the semicrystalline materials. (E-EP)6 showed minimal improvements in the final set compared to (E-EP)2, which is dictated by plastic deformation of the crystallites. On the other hand, the final recovery of (E-C-EP)6 was higher than any of the other semicrystalline TPEs, despite the fact that the composite hard domains or the star architecture alone failed to increase the recovery. In comparison with other branched architectures reported in the literature, these findings underscore the utility of using well-defined star block copolymers to enhance strain hardening and increase recovery in block polymer TPEs.

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.6b02175. Gel permeation chromatography traces of 6-arm stars prior to hydrogenation, form factor calculations for (CEP)2 and (C-EP)6, description of cyclic strain test protocol and an example of the raw data, and recovery and hysteresis of Kraton D1111 compared to (C-EP)2 and (C-EP)6 (PDF) H

DOI: 10.1021/acs.macromol.6b02175 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules



(17) Bi, L.-K.; Fetters, L. J. Synthesis and Properties of Block Copolymers. 3. Polystyrene-Polydiene Star Block Copolymers. Macromolecules 1976, 9, 732−742. (18) Shim, J. S.; Kennedy, J. P. Novel Thermoplastic Elastomers. II. Properties of Star-block Copolymers of PSt-b-PIB Arms Emanating from Cyclosiloxane Cores. J. Polym. Sci., Part A: Polym. Chem. 1999, 37, 815−824. (19) Shim, J. S.; Kennedy, J. P. Novel Thermoplastic Elastomers. III. Synthesis, Characterization, and Properties of Star-block Copolymers of Poly(indene-b-isobutylene) Arms Emanating from Cyclosiloxane Cores. J. Polym. Sci., Part A: Polym. Chem. 2000, 38, 279−290. (20) Jacob, S.; Majoros, I.; Kennedy, J. P. Novel Thermoplastic Elastomers: Star-Blocks Consisting of Eight Poly(styrene-b-isobutylene) Arms Radiating from a Calix[8]arene Core. Rubber Chem. Technol. 1998, 71, 708−721. (21) Juhari, A.; Mosnácě k, J.; Yoon, J. A.; Nese, A.; Koynov, K.; Kowalewski, T.; Matyjaszewski, K. Star-like Poly(n-butyl acrylate)-bpoly(α-methylene-γ-butyrolactone) Block Copolymers for High Temperature Thermoplastic Elastomers Applications. Polymer 2010, 51, 4806−4813. (22) Nese, A.; Mosnácĕ k, J.; Juhari, A.; Yoon, J. A.; Koynov, K.; Kowalewski, T.; Matyjaszewski, K. Synthesis, Characterization, and Properties of Starlike Poly(n-butyl acrylate)-b-poly(methyl methacrylate) Block Copolymers. Macromolecules 2010, 43, 1227−1235. (23) Kwee, T.; Taylor, S. J.; Mauritz, K. A.; Storey, R. F. Morphology and Mechanical and Dynamic Mechanical Properties of Linear and Star Poly(styrene-b-isobutylene-b-styrene) Block Copolymers. Polymer 2005, 46, 4480−4491. (24) Jeusette, M.; Leclère, P.; Lazzaroni, R.; Simal, F.; Vaneecke, J.; Lardot, T.; Roose, P. New “All-Acrylate” Block Copolymers: Synthesis and Influence of the Architecture on the Morphology and the Mechanical Properties. Macromolecules 2007, 40, 1055−1065. (25) Imaeda, T.; Hashimoto, T.; Irie, S.; Urushisaki, M.; Sakaguchi, T. Synthesis of ABA-Triblock and Star-Diblock Copolymers with Poly(2-adamantyl vinyl ether) and Poly(n-butyl vinyl ether) Segments: New Thermoplastic Elastomers Composed Solely of Poly(vinyl ether) Backbones. J. Polym. Sci., Part A: Polym. Chem. 2013, 51, 1796−1807. (26) Storey, R. F.; Chisholm, B. J.; Choate, K. R. Synthesis and Characterization of PS-PIB-PS Triblock Copolymers. J. Macromol. Sci., Part A: Pure Appl.Chem. 1994, 31, 969−987. (27) Pakula, T.; Koynov, K.; Boerner, H.; Huang, J.; Lee, H.-i.; Pietrasik, J.; Sumerlin, B.; Matyjaszewski, K. Effect of Chain Topology on the Self-organization and the Mechanical Properties of Poly(n-butyl acrylate)-b-polystyrene Block Copolymers. Polymer 2011, 52, 2576− 2583. (28) Asthana, S.; Kennedy, J. P. Star−block Polymers of Multiple Polystyrene-b-Polyisobutylene Arms Radiating from a Polydivinylbenzene Core. J. Polym. Sci., Part A: Polym. Chem. 1999, 37, 2235−2243. (29) Hsieh, H. L.; Quirk, R. P. Applications of Styrenic Thermoplastic Rubbers in Plastics Modifications, Adhesives, and Footwears. In Anionic Polymerization: Principles and Practical Applications; Marcel Dekker, Inc.: New York, 1996; pp 533−585. (30) Kong, J. F.; Lipik, V.; Abadie, M. J. M.; Deen, G. R.; Venkatraman, S. S. Characterization and Degradation of Elastomeric Four-Armed Star Copolymers Based on Caprolactone and L-Lactide. J. Biomed. Mater. Res., Part A 2012, 100A, 3436−3445. (31) Kong, J. F.; Lipik, V.; Abadie, M. J. M.; Deen, G. R.; Venkatraman, S. S. Biodegradable Elastomers Based on ABA Triblocks: Influence of End-Block Crystallinity on Elastomeric Character. Polym. Int. 2012, 61, 43−50. (32) Burns, A. B.; Register, R. A. Strategies for the Synthesis of WellDefined Star Polymers by Anionic Polymerization with Chlorosilane Coupling and Preservation of the Star Architecture during Catalytic Hydrogenation. Macromolecules 2016, 49, 2063−2070. (33) Register, R. A.; Bell, T. R. Miscible Blends of Zinc-neutralized Sulfonated Polystyrene and Poly(2,6-dimethyl 1,4-phenylene oxide). J. Polym. Sci., Part B: Polym. Phys. 1992, 30, 569−575.

AUTHOR INFORMATION

Corresponding Author

*Tel +1 609 258 4691; fax +1 609 258 0211; e-mail register@ princeton.edu (R.A.R.). ORCID

Richard A. Register: 0000-0002-5223-4306 Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This work was generously supported by the National Science Foundation, Polymers Program (DMR-1402180).



REFERENCES

(1) Holden, G.; Legge, N. R.; Quirk, R. P.; Schroeder, H. E. Thermoplastic Elastomers, 2nd ed.; Hanser/Gardner Publications: Cincinnati, OH, 1996. (2) Holden, G.; Bishop, E. T.; Legge, N. R. Thermoplastic Elastomers. J. Polym. Sci., Part C: Polym. Symp. 1969, 26, 37−57. (3) Myers, S. B.; Register, R. A. Extensibility and Recovery in a Crystalline-Rubbery-Crystalline Triblock Copolymer. Macromolecules 2009, 42, 6665−6670. (4) Bishop, J. P.; Register, R. A. Thermoplastic Elastomers with Composite Crystalline-Glassy Hard Domains and Single-Phase Melts. Macromolecules 2010, 43, 4954−4960. (5) Burns, A. B.; Register, R. A. Thermoplastic Elastomers via Combined Crystallization and Vitrification from Homogeneous Melts. Macromolecules 2016, 49, 269−279. (6) Schultz, J. Mechanical Behavior of Semicrystalline Polymers. In Polymer Materials Science; Prentice-Hall, Inc.: Englewood Cliffs, NJ, 1974; pp 466−511. (7) Hiss, R.; Hobeika, S.; Lynn, C.; Strobl, G. Network Stretching, Slip Processes, and Fragmentation of Crystallites during Uniaxial Drawing of Polyethylene and Related Copolymers. A Comparative Study. Macromolecules 1999, 32, 4390−4403. (8) Nitta, K.-h.; Takayanagi, M. Novel Proposal of Lamellar Clustering Process for Elucidation of Tensile Yield Behavior of Linear Polyethylenes. J. Macromol. Sci., Part B: Phys. 2003, 42, 107−126. (9) Men, Y.; Rieger, J.; Strobl, G. Role of the Entangled Amorphous Network in Tensile Deformation of Semicrystalline Polymers. Phys. Rev. Lett. 2003, 91, 095502. (10) López-Barrón, C. R.; Eberle, A. P. R.; Yakovlev, S.; Bons, A.-J. Structural Origins of Mechanical Properties and Hysteresis in SIS Triblock Copolymers/Polystyrene Blends with Spherical Morphology. Rheol. Acta 2016, 55, 103−116. (11) Pakula, T.; Saijo, K.; Kawai, H.; Hashimoto, T. Deformation Behavior of Styrene-Butadiene-Styrene Triblock Copolymer with Cylindrical Morphology. Macromolecules 1985, 18, 1294−1302. (12) Séguéla, R.; Prud’homme, J. Affinity of Grain Deformation in Mesomorphic Block Polymers Submitted to Simple Elongation. Macromolecules 1988, 21, 635−643. (13) Honeker, C. C.; Thomas, E. L. Impact of Morphological Orientation in Determining Mechanical Properties in Triblock Copolymer Systems. Chem. Mater. 1996, 8, 1702−1714. (14) Daniel, C.; Hamley, I. W.; Mortensen, K. Effect of Planar Extension on the Structure and Mechanical Properties of Polystyrene− Poly(ethylene-co-butylene)−Polystyrene Triblock Copolymers. Polymer 2000, 41, 9239−9247. (15) Indukuri, K. K.; Lesser, A. J. Comparative Deformational Characteristics of Poly(styrene-b-ethylene-co-butylene-b-styrene) Thermoplastic Elastomers and Cross-linked Natural Rubber. Polymer 2005, 46, 7218−7229. (16) Hsieh, H. L.; Quirk, R. P. Styrenic Thermoplastic Elastomers. In Anionic Polymerization: Principles and Practical Applications; Marcel Dekker, Inc.: New York, 1996; pp 475−531. I

DOI: 10.1021/acs.macromol.6b02175 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules (34) Porod, G. General Theory. In Small Angle X-ray Scattering, Glatter, O., Kratky, O., Eds.; Academic Press Inc.: London, 1982; pp 17−51. (35) Wunderlich, B.; Dole, M. Specific Heat of Synthetic High Polymers. VIII. Low Pressure Polyethylene. J. Polym. Sci. 1957, 24, 201−213. (36) Matsen, M. W.; Schick, M. Microphase Separation in Starblock Copolymer Melts. Macromolecules 1994, 27, 6761−6767. (37) Fetters, L. J.; Lohse, D. J.; Richter, D.; Witten, T. A.; Zirkel, A. Connection between Polymer Molecular Weight, Density, Chain Dimensions, and Melt Viscoelastic Properties. Macromolecules 1994, 27, 4639−4647. (38) Stasiak, J.; Squires, A. M.; Castelletto, V.; Hamley, I. W.; Moggridge, G. D. Effect of Stretching on the Structure of Cylinderand Sphere-Forming Styrene−Isoprene−Styrene Block Copolymers. Macromolecules 2009, 42, 5256−5265. (39) Rangarajan, P.; Register, R. A.; Fetters, L. J. Morphology of Semicrystalline Block Copolymers of Ethylene-(ethylene-alt-propylene). Macromolecules 1993, 26, 4640−4645. (40) Koo, C. M.; Wu, L.; Lim, L. S.; Mahanthappa, M. K.; Hillmyer, M. A.; Bates, F. S. Microstructure and Mechanical Properties of Semicrystalline-Rubbery-Semicrystalline Triblock Copolymers. Macromolecules 2005, 38, 6090−6098. (41) Olvera de la Cruz, M.; Sanchez, I. C. Theory of Microphase Separation in Graft and Star Copolymers. Macromolecules 1986, 19, 2501−2508. (42) Benoît, H.; Hadziioannou, G. Scattering Theory and Properties of Block Copolymers with Various Architectures in the Homogeneous Bulk State. Macromolecules 1988, 21, 1449−1464. (43) Hashimoto, T.; Ijichi, Y.; Fetters, L. J. Order−disorder Transition of Starblock Copolymers. J. Chem. Phys. 1988, 89, 2463− 2472. (44) Ijichi, Y.; Hashimoto, T.; Fetters, L. J. Order-disorder Transition of Star-block Copolymers. 2. Effect of Arm Number. Macromolecules 1989, 22, 2817−2824. (45) Kar, K. K.; Bhowmick, A. K. High-Strain Hysteresis of Rubber Vulcanizates Over a Range of Compositions, Rates, and Temperatures. J. Appl. Polym. Sci. 1997, 65, 1429−1439. (46) Mohajer, Y.; Wilkes, G. L.; Wang, I. C.; McGrath, J. E. Hydrogenated Linear Block Copolymers of Butadiene and Isoprene: Effects of Variation of Composition and Sequence Architecture on Properties. Polymer 1982, 23, 1523−1535. (47) Sato, K. The Mechanical Properties of Filled Polymers. Prog. Org. Coat. 1976, 4, 271−302. (48) Diamant, J.; Williams, M. C.; Soane, D. S. Microstructural Diagnosis of Block Copolymer Nonlinear Mechanical Properties I. Uniaxial Stress/Strain. Polym. Eng. Sci. 1988, 28, 207−220. (49) Quirk, R. P.; Morton, M. Research on Anionic Triblock Copolymers. In Thermoplastic Elastomers; Holden, G., Legge, N. R., Quirk, R. P., Schroeder, H. E., Eds.; Hanser/Gardner Publications: Cincinnati, OH, 1996; pp 71−100. (50) Séguéla, R.; Prud’homme, J. Structural and Mechanical Properties of a Polyethylene-Based Thermoplastic Elastomer. Polymer 1989, 30, 1446−1455. (51) Staudinger, U.; Weidisch, R.; Zhu, Y.; Gido, S. P.; Uhrig, D.; Mays, J. W.; Iatrou, H.; Hadjichristidis, N. Mechanical Properties and Hysteresis Behaviour of Multigraft Copolymers. Macromol. Symp. 2006, 233, 42−50. (52) Hotta, A.; Cochran, E.; Ruokolainen, J.; Khanna, V.; Fredrickson, G. H.; Kramer, E. J.; Shin, Y.-W.; Shimizu, F.; Cherian, A. E.; Hustad, P. D.; Rose, J. M.; Coates, G. W. Semicrystalline Thermoplastic Elastomeric Polyolefins: Advances Through Catalyst Development and Macromolecular Design. Proc. Natl. Acad. Sci. U. S. A. 2006, 103, 15327−15332. (53) O’Connor, K. S.; Watts, A.; Vaidya, T.; LaPointe, A. M.; Hillmyer, M. A.; Coates, G. W. Controlled Chain Walking for the Synthesis of Thermoplastic Polyolefin Elastomers: Synthesis, Structure, and Properties. Macromolecules 2016, 49, 6743−6751.

(54) Ohtaki, H.; Deplace, F.; Vo, G. D.; LaPointe, A. M.; Shimizu, F.; Sugano, T.; Kramer, E. J.; Fredrickson, G. H.; Coates, G. W. AllylTerminated Polypropylene Macromonomers: A Route to Polyolefin Elastomers with Excellent Elastic Behavior. Macromolecules 2015, 48, 7489−7494.

J

DOI: 10.1021/acs.macromol.6b02175 Macromolecules XXXX, XXX, XXX−XXX