Two Distinct Morphologies for Semicrystalline Isotactic Polypropylene

Jun 20, 2018 - Above a critical shear rate (inverse of long-chain relaxation time 1/τ), the onset .... After shear, we monitor crystallization in the...
2 downloads 0 Views 14MB Size
Download PDF
Article Cite This: Macromolecules XXXX, XXX, XXX−XXX

Two Distinct Morphologies for Semicrystalline Isotactic Polypropylene Crystallized after Shear Flow Behzad Nazari,† Han Tran,‡ Burke Beauregard,‡ Matthew Flynn-Hepford,† Douglas Harrell,§ Scott T. Milner,‡ and Ralph H. Colby*,† Materials Science and Engineering and the Materials Research Institute and ‡Department of Chemical Engineering, Penn State University, University Park, Pennsylvania 16802, United States § The Phillips 66 Company, Linden, New Jersey 07036, United States Downloaded via KAOHSIUNG MEDICAL UNIV on June 21, 2018 at 01:56:03 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

†

S Supporting Information *

ABSTRACT: Application of shear flow to molten, highly isotactic polypropylene (iPP) results in two different morphology transitions: (1) above a certain shear rate but below a critical shear stress σ*, flow-induced precursors nucleate many small crystallites; (2) for shear stress above σ*, shish precursors nucleate highly oriented shish-kebab morphology. Herein we study flow-induced crystallization (FIC) in iPP with different molecular weights, using rotational and capillary rheometry. Since precursors created by shear are quite stable, we can also use differential scanning calorimetry (DSC) and polarized optical microscopy (POM) to study crystallization, melting, and morphology of iPP samples with different shear histories. Above a critical shear rate (inverse of long-chain relaxation time 1/τ), the onset of crystallization on cooling shifts to higher temperatures compared to unsheared samples. POM micrographs see a clear border between the regions affected by FIC (with γ̇ > 1/τ) and regions crystallizing as though they had not been sheared. FIC results in much smaller crystallites, so-called rice grains of order 1 μm in size. Above a critical shear stress (σ* ∼ 0.11 MPa) in the rotational rheometer, the morphology transitions to a shish-kebab structure. Shish appear in micrographs as highly aligned birefringent regions; in DSC, flow-induced shish further accelerate the onset of crystallization. In the rheometer, sheared samples with σ < σ* at 170 °C (above Tm) behave as a viscoelastic liquid identical to unsheared samples, whereas strongly sheared samples with σ > σ* behave as weak gels, revealing the presence of a percolating network of shish. In capillary rheometry, samples sheared above this threshold stress likewise show an abrupt increase in apparent viscosity.

■

INTRODUCTION Flow fields are inevitable components of polymer processing. Under such conditions, conformational changes and order can develop to an extent far from quiescence. Within the polymer science community, it has become clear that flow-induced conformational ordering has a significant influence on crystallization rate and semicrystalline morphology; depending on flow strength, crystallization kinetics can be enhanced by orders of magnitude1−4 and crystal morphology can be modified from large spherulites to orders of magnitude smaller crystallites5−9 and eventually transform into oriented crystallites (shish-kebab).10−17 These flow-induced modifications in the polymer crystallinity and crystalline morphology are intimately associated with the mechanical properties of the final product, hence attracting a great deal of attention in both industrial and academic settings.18−20 Rheometers are widely used to study flow-induced crystallization (FIC), since these instruments enable researchers to apply flow at elevated temperatures (>Tm, the polymer nominal melting point) and to monitor crystallization via oscillatory time sweeps upon lowering temperature ( τ , with τR the Rouse time of

are robust to annealing. This enables us to use DSC on previously sheared rheometer samples to examine the recrystallization exotherm on cooling. Finally, we can observe the final morphology of samples from the rheometer directly with POM. Combining these techniques reveals dramatic differences between quiescently crystallized iPP and samples containing FIC precursors or shish. We can also detect the formation of shish before cooling the sample directly in the rheometer by their effect on melt viscoelasticity. As shish form, they give rise to a percolating elastic network, detectable as a weak gel in linear viscoelasticity above Tm in the rheometer. Experiments in a capillary rheometer detect a corresponding jump in apparent viscosity when the shear stress in most of the sample exceed the critical threshold σ*.

R

the longest chains in the melt).4,8,9,31−33 Applying such a shear rate to the molten polymer is known to form precursors that promote nucleation, resulting in faster crystallization and smaller crystallites.4,8,9,34,35 In other words, a small fraction of very high molecular weight molecules can enhance FIC for a given polymer melt. The work done by flow per unit volume of melt, called specific work (defined as for constant shear rate γ̇, total strain γ, and viscosity η), has been found to be a key parameter controlling the strength of FIC.2,4,8,9,17,27−30 It has been shown that as the applied work increases, the crystallite number density increases and oriented crystalline structures (“rice grains“) become more prevalent. At the same time, larger spherulites vanish from the semicrystalline morphology,9,36 simply because they would form too slowly compared to the rice grains. This change in morphology was found to be associated with accelerated crystallization for samples with higher specific work. It is not clear yet how specific work leads to smaller crystalline structures and faster crystallization, but the initial structure of shear-induced precursors may give important clues. Time-resolved small-angle X-ray scattering (SAXS) and Fourier transform infrared spectroscopy experiments have been interpreted as showing long-range density fluctuations, reflecting close packing between neighboring chains. These fluctuations have been emphasized in previous work on sheared polymer melts, which suggest flow-induced precursors consist of multiple chain stems rather than a single chain segment.37,38 It has been speculated that during a long interval of shear at high stress (hence large specific work), stretched chains get together to form the precursors of FIC.36 During flow at high stress, long chains become highly stretched away from random coil conformations, which can promote close lateral packing between chains and form dense, long fibrillar bundles of molecules, or shish.10−17 The shish (as crystallization precursors) can rapidly crystallize and subsequently become overgrown by lamellar crystals (kebabs).17,39,40 Using pressuredriven flow for iPP, Balzano et al.17 reported that the size and morphology of the formed crystallites strongly depend on shear stress. Not varying the iPP MW, the authors revealed as the shear stress increased, the spherulite size decreased until a critical shear stress σ* ∼ 0.14 MPa, where the morphology suddenly transformed into highly birefringent shish-kebab structures stretched along the flow direction. In this work, using a rotational rheometer capable of shearing at both very low shear rates and at high enough rates to form shish structures, we carry out experiments to establish criteria for the onsets of (1) FIC and (2) shish formation in iPPs with different MW and MWD. After shear, we monitor crystallization in the rheometer with fixed frequency oscillatory shear while cooling. In this way, we examine the effect of shear history on nonisothermal crystallization of iPP. Previous studies have found that once formed, flow-induced precursors

■

EXPERIMENTAL SECTION

Rotational rheometry was conducted using an ARES-G2 from TA Instruments. Based on the type of intended tests, parallel plate geometries (diameter: 8 and 25 mm; gap: 0.05−0.5 mm) as well as a cone−plate geometry (diameter: 25 mm; truncation: 52 μm; cone angle: 1°) were used in this work. This rotational rheometer uses a heated stream of nitrogen for temperature control to help prevent thermal degradation. To minimize thermal/shear history in the iPP pellets, each sample was first annealed at 220 °C for ∼10 min. All the sheared disks were carefully brought to room temperature in nitrogen prior to removal for POM/DSC experiments. Transmission POM used an Olympus BX51 microscope equipped with a hot stage (HCS621 V from Instec, Inc. with a nitrogen blanket). Disks made in the rheometer for POM were sufficiently thin (∼50 μm) to ensure easy transmission of light through the sample for clear images to be captured by the objective lens (4×, 10×, and 20× magnification). DSC was run using a Q2000-DSC from TA Instruments with samples protected via a constant nitrogen purge of 50 mL/min. Samples for DSC were prepared using the rheometer, with a sample thickness of about 0.5 mm, so as to maintain a constant sample mass in the DSC pans (3−6 mg). The DSC pans used were aluminum hermetic with lids pressed onto the top. Importantly, to prevent torque/axial force overload while preparing sheared samples in the rheometer at high applied stresses for POM and DSC studies, the transducer was locked before applying shear. A Rosand dual barrel capillary rheometer (RH10 from Malvern Instruments) was used for studying rheology in pressure-driven flow, with different capillary dies: diameter 0.5, 0.75, and 1.0 mm and length 10, 20, and 30 mm, with a nitrogen blanket surrounding the barrel entrance. Die diameter and length were chosen based on the viscosity of the iPP samples; i.e., samples with the lowest MW were run on the die with diameter 0.5 mm and length 30 mm whereas samples with highest MW were pushed through the die with diameter 1.0 mm and length 10 mm. This helped to prevent transducer overload at high piston speeds (high shear stress at the wall). Bagley correction was taken into account by subtracting the entrance pressure drop (measured using an orifice die with almost zero length, 0.25 mm) from the pressure measured while running the tests, following Cogswell.41 Capillary rheometry was conducted at 170 °C after erasing history at 220 °C for ∼10 min. Six commercial iPP samples (iPP1, 2, 3, 4, 5, and 6) were used from different sources. In our previous work,8 MW and size-exclusion chromatography (SEC) MWD were reported for iPP1, 2, 3, 4, and 5 (Table 1). We used this MWD (see Figure 5 of ref 8) to estimate the high-MW tails of iPP1, 2, 3, 4, and 5, defining a “maximum” molecular weight (Mmax) by the exponential decay of the high-Mw tail in the distribution (see Table 1). Specifically, the natural log of the SEC chain count was plotted vs MW, and the slope of the line at large MW was defined as −1/Mmax. Hamad et al.8 performed calculations of the LVE response using the BoB (branch on branch) program to determine rheological parameters: entanglement molecular weight Me = 5250 g/mol and entanglement strand Rouse time τe = 1.5 × 10−7 s at 170 °C.4,8,42 B

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules Table 1. Linear Viscoelastic (at 170 °C) and Molecular Characteristics of Six Polypropylenes iPP# MW (kg/mol)a η0 (Pa·s) τ (s) m=1−n Mmax (kg/mol) Mmax/Me τRmax (s)c XS (%)d

1

2

3

4

5

6

448

158

236

462

474

627b

27900 2.64 0.58 16000

1190 0.05 0.59 2500

5780 0.43 0.55 11000

32700 1.61 0.58 27000

77000 6.41 0.56 21000

145000 9.78 0.62

3000 1.39

760 0.09

2100 0.66

5300 4.14

3900 2.38

3.3

1.9

2.2

1.7

4.8

The Cross model43 (η =

■

RESULTS To measure quiescent nonisothermal crystallization, iPP samples were subjected to cooling scans at 5 °C/min in the DSC after annealing at 220 °C for 10 min (Figure 2a). The DSC data were integrated to obtain the percent crystallinity (Xc%) versus temperature. The percent crystallinity is defined as the fractional enthalpy ΔH/ΔH0 × 100, where ΔH is the crystallization energy released from the melt during cooling and ΔH0 = 209 J/g is the enthalpy of fusion for iPP crystals.44 The onset of crystallization (Toc) upon cooling from the melt is an indication of nucleation and overall crystallization rate.45 It has been reported that lower MW polyolefin samples crystallize faster than the higher MW ones during cooling from the melt.46,47 This was related to the higher molecular mobility at low MW, making it easier for the polymer chains to be extricated from the entangled melt. This explains the trend in the onset of crystallization of Figure 2a for iPP1, 2, 3, 4, and 5: Toc for iPP2 > iPP3 > iPP4 > iPP1 > iPP5, which is consistent with MW for iPP2 < iPP3 < iPP4 < iPP1 < iPP5. Interestingly, iPP6, with the highest MW, was found to begin crystallization at a higher temperature than any of the other samples (Toc ∼ 128 °C). This discrepancy might be attributed to the larger amount of catalyst residues left in iPP6 after polymerization (confirmed by the manufacturer), which may account for its ease of crystallizing at high temperatures, due to the heterogeneous nucleation effect of the mineral particles.45,48,49 One can also argue that perhaps the thermal/shear history in iPP6 is not sufficiently erased in 10 min at 220 °C (due to its high MW and slower chain relaxation),4 boosting crystallization via flowinduced nucleation. Indeed, following ref 4, iPP6 was annealed for 5 h at 250 °C to remove the FIC effects that were apparently in iPP6 from prior extrusion (see Figure 3 and later the caption of Figure 11e). POM was used to directly image the formation of spherulites in quiescently crystallizing iPPs (Figure 3) and to clarify and complement the DSC results (Figure 2). Prior to crystallizing under POM at 135 °C, the samples were brought to 220 °C and kept for 10 min in an effort to erase their thermal history.

3.6

Weight-average molecular weight reported in refs 4 and 8 for iPP1, 2, 3, 4, and 5. bEstimated from η0 = 8.5 × 10−5MW3.3 at 170 °C. cRouse time of the longest chains at 170 °C, estimated from 2

( ). M max Me

d

Xylene-soluble fraction, indicative of tacticity

(ASTM D5492). 2

( ) , the Rouse relaxation time of the longest

Using τR max = τe

M max Me

) was fit to the data in Figure 1

for each iPP to calculate zero-shear viscosity (η0), relaxation time (τ), and shear-thinning exponent (m = 1 − n where n is the power law index), listed in Table 1. The inset in Figure 1 shows the log−log plot of zero-shear viscosity vs MW for iPP1, 2, 3, 4, and 5 (η0 ∼ MW3.3). This plot is used to estimate MW for iPP6 (∼627 kg/mol). The determination of the xylene-soluble fraction (XS) in the iPPs was used to estimate the atactic content; the results are shown in Table 1 and are consistent with our published work.8

a

τR max = τe

η0 1 + (τγ )̇ m

chains was estimated, revealed in Table 1. Figure 1 depicts the complex viscosity for iPP1, 2, 3, 4, 5, and 6 obtained via frequency sweeps (at 170 °C and 0.01 strain amplitude).

Figure 1. Complex viscosity (obtained from frequency sweep at strain amplitude 0.01) as a function of frequency for different iPPs at 170 °C. Curves are fits to the Cross model43 with parameters listed in Table 1. The inset plots zero-shear viscosity vs MW for iPP1, 2, 3, 4, and 5 at 170 °C.

Figure 2. DSC scans for the six quiescent iPPs: (a) apparent crystallinity (Xc%) as a function of temperature for different iPPs at a cooling rate of 5 °C/min after annealing at 220 °C for 10 min, followed by (b) heating scans (10 °C/min). The red dotted curves are for iPP6 as received while the black dotted curves are iPP6 after annealing at 250 °C for 5 h to remove FIC effects (similar annealing of the other five iPP samples had no effect on crystallization or melting). C

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 4. Schematic of a sheared disk: (a) regions on the disk with different crystal morphology, (b) description of the disk dimensions, and (c) a sample micrograph (1.42 × 1.07 mm captured using a 10× objective lens) of an iPP disk sheared at 170 °C followed by crystallization at 135 °C. The pink and purple dashed lines indicate the no-FIC/FIC and FIC/shish borders, respectively. The intermediate FIC region is darkest because micrometer-size crystallites scatter the most light. Figure 3. Quiescent crystalline morphology of iPP samples captured using POM at 135 °C after annealing at 220 °C for ∼10 min. Each window is 1.42 × 1.07 mm, using a 10× objective lens. Average spherulite diameter: 300, 50, 60, 110, 250, and 35 μm for iPP1, 2, 3, 4, 5, and 6, respectively.

rΩ

radial position (r), γ(̇ r ) = h , where h is the disk thickness. Using the Cross model (fits shown in Figure 1, parameters in Table 1), the shear viscosity at a given radius can be estimated η as η(r ) = 1 + (τγ0 (̇ r))n . The shear stress σ(r) can be calculated as

As shown in Figure 3, for all materials except iPP2, the quiescently crystallized iPPs consist of spherulites rather uniform in size. In contrast, iPP2 seems to have a bimodal distribution of spherulite diameters, consistent with the multiple melting peaks of crystalline iPP2 observed in DSC heating scans (Figure 2b). The average spherulite diameter varied among our different iPP samples, with the same trend as observed in the crystallization onset temperature Toc. The samples that crystallized most readily (at the highest Toc) have the smallest spherulites. From smallest to largest, the diameters are iPP6 < iPP2 < iPP3 < iPP4 < iPP1 < iPP5. The very tiny spherulites of iPP6 formed rapidly when the sample was brought to the crystallization temperature (135 °C) in POM. In most commercial polymers, primary nucleation is commonly produced by heterogeneous nuclei, including catalyst residues and other impurities of unknown origin.47,48 The very small spherulite diameter for iPP6 compared to those of iPP1, 4, and 5 (which form crystals very slowly, consistent with their high MW) can be attributed to the higher level of catalyst residue in iPP6 compared to the other samples, combined with residual shear history in iPP6 not erased by annealing for 10 min at 220 °C (see discussion of Figure 2 earlier and the caption of Figure 11e later). After characterizing quiescent crystallization of our iPP samples, we investigated FIC by shearing iPP disks in a parallel-plate rheometer (Figure 4). Before each test, the sample was kept at 220 °C for about 10 min to minimize effects of prior thermal/shear history for the pelletized materials. Following our past work,4,8,9 all shearing was performed at 170 °C. The shear rate (γ̇) applied to the sample (Figure 4b) is linear in both angular velocity (Ω) and

σ(r) = γ̇(r)η(r). We can then calculate specific work for a given shearing time as W(r) = σ(r)γ̇(r)t at any radial position r. When a disk of iPP is sheared, specific regions with different crystal morphology on cooling form at different radial positions on the disk.36,40 These regions are sketched in Figure 4a and visualized in Figure 4c. Near the disk center, the flow is weak and the shear rate, stress, and work are all small, so that chains are not much aligned or stretched to promote nucleation. As a result, the center crystallizes in a quiescent way (the no-FIC region, r < rf). Beyond rf, the next region is FIC (rf < r < rs); here, the flow is sufficient to stretch the longest chains and promote nucleation. Crystallites in the FIC region are significantly smaller than those in the no-FIC region. Micrographs are unable to resolve the micrometer-size crystallites in the FIC region, but Hamad et al.9 used AFM to image such structures and observed “rice grain” domains roughly 3 μm in length and 1 μm in diameter for iPP1. At still larger radial positions r > rs, the shear stress becomes high enough to produce a shish-kebab morphology that is easily detected by very strong birefringence. The outer edge of the sample beyond the shish region always shows some edge fracture, especially at high γ̇. Throughout this paper, results corresponding to the quiescent, FIC, and shish regions are depicted in green, yellow, and blue, respectively. Different measures of the strength of flow have been proposed as important parameters governing FIC: shear rate, shear stress, specific work. Here, we design experiments to separate the effects of these parameters. To examine the effect of specific work on crystallization rate, we kept the shear rate at the disk edge constant and sheared for different amounts of time. In all cases, samples were held at 220 °C for about 10 D

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules min and sheared at 170 °C. The shear rate at the perimeter was fixed at γ̇ = 10 s−1, larger than the reciprocal of the Rouse time of the longest chains, except for iPP2 which was sheared at 80 s−1 to reach a perimeter Weissenberg number of 4. After shearing, each sample was cooled at 1 °C/min, and crystallization monitored with oscillatory shear at 0.5 rad/s and 0.01 strain amplitude. The temperature at which the modulus crossover (tan δ = 1) occurs is taken as the crystallization temperature. (This protocol of cooling after shear and observing the crystallization temperature is simpler than our previous approach of quenching after shear and noting the crystallization time because it avoids the need to quickly quench and stabilize the temperature.) Representative results for dynamic response at 0.5 rad/s versus temperature on cooling are given in the Supporting Information (Figure S1). Figure 5 shows the crystallization temperature versus specific work for each of our iPP samples. For all samples except iPP2,

While the identity of the FIC precursors is not yet proven, one prevailing view37,38 is that precursors contain multiple stems as opposed to a single isolated stretched chain segment. Such multistems require time to assemble and hence require larger W. iPP6 differs from the other samples showing FIC, in that the crystallization temperature continues to slowly increase as specific work (shearing time) is increased beyond 15 MPa. iPP6 also differs from the other samples in that it contains particularly high levels of catalyst residue nanoparticles, which are thought to act as heterogeneous nucleating agents.50,51 These two observations support the hypothesis that stretched long chains adsorb onto nanosized solid particles in the polymer melt to produce FIC precursors.9 Adsorption of stretched chains onto particles has also been suggested to account for the large activation energy associated with the very slow disappearance of the FIC precursors at elevated temperatures, after cessation of shear.9 By this view, the saturation of FIC effects beyond ∼10 MPa of specific work for iPP1, 3, 4, and 5 is a result of those samples running out of either long chains to stretch and adsorb to nanoparticle impurities or nanoparticles to adsorb to. For iPP6, which has more of both than all the other samples, further increases in applied work (hence continued shearing) apparently can produce additional nuclei and further increase the crystallization temperature.51 Monitoring crystallization in the rheometer is effective, but we cannot tell where crystallization is occurring or how regions of the disk with different shear rates respond to the flow. Imaging the sheared and crystallized disk with POM provides a useful complementary information. Figure 6 shows POM

Figure 5. Crystallization temperature as a function of specific work applied at 10 s−1 at the disk edge (for iPP2, the perimeter shear rate was much higher: 80 s−1, iPP2 still shows no FIC effect on crystallization temperature) at 170 °C, by measuring the moduli crossover temperature using an oscillatory shear temperature sweep: at 1 °C/min cooling rate, 0.01 strain amplitude, and 0.5 rad/s frequency.

FIC effects increase sharply from 0 to 7 MPa and tend to saturate at around W = 15−20 MPa. Similar results were observed for different shear rates. These results are consistent with our previous findings4,8,9 for the effect of W on isothermal crystallization of iPP: iPP2 never shows any FIC, and iPP1, 3, 4, and 5 each exhibited similar saturated work levels. In that work, the protocol was shearing at 170 °C followed by a quench to 141 °C without undershooting, where small amplitude oscillatory shear at 0.5 rad/s was used to monitor crystallization. In contrast, iPP2 did not display any FIC; its crystallization temperature was around 125 °C, regardless of the shear history applied to the sample. We conclude that iPP2 does not have long enough chains to create FIC precursors. This hypothesis was tested by applying higher shear rates (>500 s−1) to iPP2, sufficient to stretch the longest chains in the sample. Even these high shear rates did not increase the crystallization temperature. Shear does make smaller (and more) spherulites for iPP2 as shear rate is increased (see Figure S2) but does not make the morphology transition to rice grains and does not show FIC effects, consistent with ref 8.

Figure 6. Crystal morphologies showing the effect of work on sheared iPP3 disks. Each window is 3 × 3 mm, using a 4× objective lens. Sheared at 170 °C in parallel plate geometry with γ̇ = 2.31 s−1 at r = 2 mm, the FIC/no-FIC border, denoted by the dashed curve. Crystallized at 135 °C.

images of disks sheared with varying specific work. For iPP3, the shear rate was kept constant (γ̇ = 2.31 s−1 at r = 2 mm), and the shearing time was tuned to adjust the specific work at r = 2 mm. The concentration of crystallites in the FIC region increases with increasing specific work. Consistent with our nonisothermal crystallization results (Figure 5), the increase in concentration of crystallites in Figure 6 appears to saturate for W > 10 MPa. Similar results for different iPPs are included in the Supporting Information (Figures S3−S5). E

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules To isolate the effect of shear rate on FIC, we take advantage of the saturation of FIC effects versus specific work. Because our iPP samples show nearly constant crystallization temperature for specific work W above 15 MPa, we fix W = 30 MPa and focus on the effect of shear rate on the resulting crystal morphology. Figure 7 shows micrographs of iPP1−6 near the

Figure 8. Critical shear rate γ̇c for the onset of FIC (from Figure 7) as a function of the molecular weight of the longest chains Mmax, from Table 1, for iPP1, 3, 4, and 5. The blue dashed line is the reciprocal of the Rouse time of the longest chains, while the red dashed line is based on the observation8 that FIC effects saturate above a shear rate of about 1 s−1, suggesting that only chains with M > 10 000 kg/mol can form FIC precursors for iPP at 170 °C. That hypothesis does account for the observation that iPP2, with Mmax = 2500 kg/mol, shows no FIC effects even up to shear rates of 500 s−1 (well above the blue dashed line). Shaded region colors are defined in Figure 4 with green = no-FIC (quiescent) and yellow = FIC. Note also that iPP3, with Mmax = 11 000 kg/mol, is close to both borders and only exhibits weak FIC (see Figure 5).

In the rheometer, we can isolate the effect of shear rate/ stress on crystallization rate by shearing at different rates and varying the shearing time to hold the specific work at the disk edge constant (W = 30 MPa). Figure 9 presents the crystallization temperature (at which tan δ = 1 at 0.5 rad/s) versus the shear stress σR at the disk edge (r = R). We also depict γ̇ on the right axis of each plot in Figure 9. For shear stress below 0.1 MPa, the crystallization temperature increases 10−30 °C (depending on the sample) above the quiescent value. Remarkably, for all of the samples that show FIC (iPP1, 3, 4, 5, and 6), a second increase of about 5−10 °C in the crystallization temperature occurs, above a stress of about 0.11 MPa. For all samples with shear stress above this threshold, the crystallized disk showed a shish-kebab structure, as evidenced by the strongly aligned and birefringent outer region in micrographs36,40 (Figure 10). The purple dotted borders on the micrographs in Figure 10 indicate the radial position on the disk where σ ≅ 0.11 MPa. This stress value is close to the 0.14 MPa that Balzano et al.17 found studying pressure-driven flow for iPP to investigate the dependency of crystalline morphology transition on shear stress. On the contrary, Mykhaylyk et al.40 used a rotational parallel disk geometry to prepare sheared disks of low-density polyethylene (LDPE). Utilizing both SAXS and POM to measure orientation, they suggested a minimum specific work 1 (constant over a wide range of shear rates with γ ̇ > τ ) serves

Figure 7. Crystalline morphology of sheared iPP parallel plate disks (sheared at 170 °C for sufficiently long periods, W > 30 MPa). The dotted line depicts the FIC/no-FIC border indicating where γ̇ > 1/τ (from Table 1). Each window is 3 × 3 mm, captured using a 4× objective lens. More examples for iPP4, 5, and 6 are found in Figures S3−S5.

radial position at which γ̇ = 1/τ (from Table 1). The micrographs show clearly that γ̇ = 1/τ is a valid criterion for the onset of FIC for a sufficiently long shearing time. This implies that FIC occurs when shear rate is large enough to stretch the longest chains. Taking averages from multiple sheared disks, the no-FIC/FIC transition occurs for iPP1, 3, 4, 5, and 6 at shear rates of 0.4, 2.2, 0.5, 0.2, and 0.1 s−1, respectively. Figure 8 displays the critical shear rate for onset of FIC versus the molecular weight of the longest chains Mmax from Table 1 for samples iPP1, 3, 4, and 5. The blue dashed line in Figure 8 is the reciprocal of the Rouse time at 170 °C: τR = 5.44 × 10−15M2 (in seconds). We note in passing that the relaxation time obtained f rom the Cross model f its in Figure 1 is essentially the same as the Rouse time of the longest chains. Furthermore, ref 8 reports a saturation of FIC effects above a shear rate of about 1 s−1. That information is displayed in Figure 8 as the red dashed vertical line, implying a threshold molecular weight for the longest chains to observe FIC: Mmax > 104 kg/mol. We note that our second lowest molecular weight sample iPP3 (black squares in Figure 5) barely shows FIC effects and is the lowest Mmax point in Figure 8, quite close to the two borders. The lowest molecular weight sample iPP2 (black circles in Figure 5) never shows any FIC even at quite high shear rates with Wi ≫ 1 and hence cannot be included in Figure 8.

R

as the critical parameter to form the shish-kebab morphology. While in iPP melts it is clear that an intermediate FIC regime creates oriented spherulites or rice grains,2,4,8,9,17,27−30 in LDPE quiescent crystallization with large spherulites transforms directly into highly birefringent shish-kebabs according to Mykhaylyk et al.36,40 Because flow-induced precursors are so robust to annealing, DSC can be used to study recrystallization of previously sheared samples as another way to study FIC. DSC samples are small, so pieces can be cut from different regions of a sheared disk to observe how the local shear history results in flowinduced precursors. F

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 9. Crystallization temperature as a function of shear stress applied at the disk edge (for iPP1, 3−6) at 170 °C (the specific work at the edge: W = 30 MPa) by measuring the moduli crossover temperature using an oscillatory shear temperature sweep: at 1 °C/min cooling rate, 0.01 strain amplitude, and 0.5 rad/s frequency. The dotted line indicates the perimeter stress at which the transition from FIC to shish occurs.

To compare material from the no-FIC, FIC, and shish regions of the disk (respectively green, yellow, and blue in Figure 4a), samples were cut from different radial positions for iPP1, 3, 4, 5, and 6. (Samples of no-FIC “quiescent crystallization” material were cut from the center of the disks.) All samples were tested with DSC cooling scans at 5 °C/min, after annealing at 175 °C for 5 min. Figure 11 displays the resulting crystallization exotherms. FIC samples have broader crystallization peaks, suggesting a superposition of nuclei formed from flow-induced precursors at higher temperatures, and nuclei that would form even in unsheared samples on cooling. In contrast, samples containing shish structures crystallize at much higher temperatures, with a crystallization peak quite distinct from that of an unsheared sample, suggesting a distinct form for the nucleating structures. For iPP4 and iPP5, the FIC peak appears only shifted with respect to the unsheared sample, suggestive of the same basic nucleation mechanism only slightly accelerated. In contrast, for

iPP1 and iPP3, the high-temperature portion of the FIC peak overlaps the crystallization peak for samples containing shish, suggesting that these FIC samples contain at least some shishlike structures. Not only can we detect the presence of shish structures by their acceleration of recrystallization in previously sheared and crystallized samples; we can also observe the formation of shish directly in the rheometer, before the sample has been cooled, by the elastic response shish impart as they percolate across the sample. To access the region of parameter space where shish form, we choose a set of steady shear rates such that the resulting steady-state stresses span the threshold for shish formation (about 0.11 MPa). Based on the Cross model fit of the complex viscosity of the iPPs (Table 1), the threshold shear rate for shish precursors (at σ ≅ 0.11 MPa) at 170 °C is about 210, 300, 65, 30, and 20 s−1 for iPP1, 3, 4, 5, and 6, respectively. G

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 10. Pictures (transmitted POM) of different iPP disks from the rheometer. The dotted border on the micrographs indicates the radial position on the disk where σ ≅ 0.11 MPa, the FIC/shish stress criterion. Each window is 1.42 × 1.07 mm captured using a 10× objective lens.

Figure 12 depicts the stress growth in the startup of shear at high enough shear rates for shish structures to form in iPP5 and iPP6. As for any entangled polymer melt, an overshoot appears before the shear stress reaches steady state.52,53 At low shear rates (with σ < 0.11 MPa at r = R), the stress plateau region either stays flat or, due to elastic instabilities and material ejection from the geometry, shear stress and normal stress start decaying. However, for sufficiently high shear stress values (σ ≥ 0.11 MPa at r = R), the normal stress begins to rise after the shear stress reaches steady state. This means the viscoelastic response of the molten polymer is becoming more gel-like, perhaps because of the formation of shish precursors. To confirm this observation of gel-like elasticity, we measured the linear viscoelastic response as a function of frequency, just before and just after applying high enough shear stress (σ > 0.11 MPa at r = R) to form shish (Figure 13). The dynamic response for the unsheared samples (squares) are typical of a polydisperse polymer melt, with the approach to terminal viscous response evident in the dynamic moduli as well as the complex viscosity, which approaches a plateau at low frequencies, η* ∼ ω0. In contrast, for the sheared samples (circles), the low-frequency response is no longer that of a viscoelastic liquid. The low-frequency dynamic moduli are larger than for the unsheared samples, and the low-frequency complex viscosity shows a power-law dependence, η* ∼ ω−0.5 for iPP5 and η* ∼ ω−0.6 for iPP6. This behavior is consistent with the formation of a percolating elastic networka weak gel, near the gel point. This effect was stronger for the high-MW

iPP6 (Figure 13b) compared to that of iPP5 (Figure 13a). For iPP1, 3, and 4, we were also able to observe this elastic indication of shish formation in the low-frequency dynamic rheology, upon applying much higher shear rates than for the higher MW iPP6 and 5. However, at these higher shear rates, some sample is ejected from the geometry, which precludes quantitative analysis. Figure 14 depicts a schematic of the network formation, as the entangled polymer melt evolves to contain shish structures above σ*. If the polymer chains are long enough and the shish density is high enough, tie chains can bridge between the shish, leading to a pronounced effect of these structures on the lowfrequency dynamic response. We observe gel-like lowfrequency deviations of magnitudes in the order iPP6 > iPP5 > iPP1 > iPP4 > iPP3, which matches the order of their relaxation times. High shear stress levels are more commonly attained experimentally in pressure-driven capillary flows. We have performed steady-shear measurements on our six iPP samples using a dual barrel capillary rheometer to look for evidence of a threshold stress for shish formation in capillary flow. In Figure S6, the raw data pressure readings in capillary and orifice dies are shown for iPP6 at different rates at 170 °C. The flow through the orifice die appears to be typical for a shear thinning material (P ∼ γ̇ 0.41). In contrast, the pressure at the entrance to the capillary die shows an abrupt increase as the shear rate is increased. H

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 11. DSC cooling scans (at 5 °C/min after annealing at 175 °C for 5 min) for different samples (a−e for iPP1, 3−6, respectively). The samples for “quiescent”, “FIC”, and “shish” were cut from the areas depicted in Figure 4a in green, yellow, and blue, respectively, far from the regime borders. For iPP6 in part e, the “quiescent” data are the as-received sample and the “annealed” data were after annealing for 5 h at 250 °C, which removed FIC effects from the iPP6 pellets that had previously been extruded. The cooling scan of the annealed iPP6 with 4 K lower peak crystallization temperature agrees well with the “quiescent” cooling scans of the other four samples, for which similar annealing did nothing, suggesting that the pellets of those samples had no FIC effect despite having been previously extruded.

Figure 12. Stress growth results at various shear rates listed for r = R (obtained at 170 °C using 8 mm parallel plate geometry after annealing the sample at 220 °C for 10 min) for (a) iPP5 which needs γ̇ = 30 s−1 to have steady-state stress 0.11 MPa and (b) iPP6 which needs γ̇ = 20 s−1 to have steady state stress 0.11 MPa.

abruptly jumps up for iPP1, 4, 5, and 6 (the four highest MW

Flow curves for iPP1, 2, 3, 4, 5, and 6 using capillary rheometry at 170 °C are shown in Figure 15. Each point represents a steady-state value (i.e., the steady-state pressure drop was stabilized over an extended period of time). The shear-thinning behavior of the molten iPPs is evident for all shear rates below σ*, whereas the steady-state shear stress

iPP samples) as the shear stress at the wall reaches about 0.27 MPa. This jump is also shown in the inset of Figure 15, which displays the apparent viscosity ( σwall ) versus shear rate. γwall ̇

I

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

Figure 15. Flow curves for the six different iPPs at 170 °C run using dual barrel capillary rheometry. The samples were annealed in the rheometer barrels at 220 °C for ∼10 min prior to the measurements at 170 °C. The slopes below 150 kPa consistently show a power law index of 0.43, a bit larger than literature data (0.33−0.36) for iPP.55−57 The inset plot shows the steady shear viscosity data as a function of shear rate. The inset cartoon schematically depicts the regions of no-FIC (green = quiescent) near the center of the die, FIC (yellow), and the dominant shish range (blue) that has σ > 0.11 MPa close enough to the wall, for σwall = 0.27 MPa. r ΔP

velocity profile, so that σ = 2L , where σ, r, ΔP, and L are shear stress, radial position, pressure drop, and capillary length, respectively.54) More than 90% of the material in capillary is outside this radius and would be expected to form shish. The cartoon in the inset of Figure 15 depicts polymer chain ordering at the conditions where the abrupt increase in the viscosity occurs due to the shish formation in about 93% of the melt passing through the capillary. Shish structures are shown in blue (similar to the cartoon in Figure 4a), formed in the extrudate in the vicinity of the capillary wall. This cartoon is consistent with prior work by Farah and Bretas,28 who used optical observations in slit dies to confirm the formation of an oriented structure near the walls, which subsequently propagated toward the center line at higher extrusion rates. The flow curve for iPP3 does not exhibit an abrupt jump in the data at a shear stress of 0.27 MPa but instead shows a significant change in the slope of the flow curve for σ > 0.27 MPa (power-law exponent switching from ∼0.46 to ∼0.77). We speculate that even though shish forms in iPP3 at sufficiently high shear stress, the chains may be too short to form enough tie chains that connect shish into the network depicted in Figure 14 (iPP3 has lower molecular weight than iPP1, 4, 5, and 6). iPP2, which has even lower molecular weight, shows a typical shear thinning behavior across the entire range in Figure 15, with no changes in slope. Recently Derakhshandeh et al.29 studied the FIC of iPP using capillary rheometry. These authors did not report abrupt jumps in flow curves but related the changes in the slope of their flow curves (similar to iPP3) to FIC effects and inferred that a “semisolid (partly crystalline) state” appeared in the polymer melt at high shear rates in the capillary. They did not quantify the shear stress at which the “semisolid state” began to form for their iPPs; but examining their flow curves, it is evident this occurred at σ ≅ 0.21 MPa, consistent with our findings. To investigate the effect of extensional flow in the die entry region, these authors used dies with different contraction angles. They found only a small effect of varying the

Figure 13. Frequency sweeps (at 170 °C and 0.01 strain amplitude) run before and after applying a shear interval with shear stress adequate for shish to form (at 170 °C using 8 mm parallel plate geometry after annealing the sample at 220 °C for 10 min): (a) 70 s−1 on iPP5 and (b) 50 s−1 on iPP6. These shear rate values resulted in steady state shear stress σ ≅ 0.12 MPa at r = R for both iPP5 and iPP6.

Figure 14. Schematic depicting network formation as the entangled polymer melt evolves to contain shish structures connected by tie chains for σ > σ*. Sufficiently long chains (blue) bridge across the shish, forming an elastic network leading to gel-like behavior in Figure 12 for sheared iPPs and the jump in shear stress shown for four capillary rheometry data sets in Figure 15.

From our experiments in the rotational rheometer, shish formation in iPP begins at a threshold stress of about 0.11 MPa. In the capillary rheometer at the threshold wall stress of 0.27 MPa, we can estimate that a local shear stress above 0.11 MPa is attained beyond a radius of 0.41R. (In making this estimate, we neglect shear-thinning and assume a Poiseuille J

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules contraction geometry, suggesting that shear deformation in the die is the main contributor to FIC in their experiments, compared to the brief interval of extensional flow. For all of our capillary dies, the entrance was flat (contraction angle 180°). From the work of Derakhshandeh et al.,29 this geometry should have negligible extensional effects on FIC, so that extrudates passing through our capillary die were mainly affected by shear flow in the die, rather than experiencing any FIC from the brief extensional flow at the die entrance.

■

AUTHOR INFORMATION

Corresponding Author

*Tel 1-814-863-3457; e-mail [email protected] (R.H.C.).

■

ORCID

Behzad Nazari: 0000-0002-9106-5445 Scott T. Milner: 0000-0002-9774-3307 Ralph H. Colby: 0000-0002-5492-6189

CONCLUSIONS Flow-induced crystallization (FIC) of various strictly linear polypropylenes with different molecular weights was studied using rotational and capillary rheometry at 170 °C, slightly above the nominal melting temperature. In the rotational rheometer, intervals of shear were applied, with the transducer locked to attain high stresses. Then cooling scans were performed, using low-frequency oscillatory shear to monitor crystallization. Above a critical shear rate, that is the inverse of Rouse relaxation time of the longest chains (1/τR), the onset of nonisothermal crystallization shifts to higher temperatures compared to unsheared material. Polarized optical microscopy (POM) micrographs of disks sheared and crystallized in a plate−plate rheometer reveal a clear border between regions affected by FIC (with γ̇ > 1/τR) and regions crystallizing quiescently, with FIC resulting in much smaller crystallites. Moreover, above a critical shear stress σc ∼ 0.11 MPa, the morphology transitions to highly aligned and birefringent shish structures. Since flow-induced nucleation precursors in iPP are quite robust to annealing, DSC observations of recrystallization can be used to compare material from different parts of a sheared and crystallized disk. Recrystallization of material from the weakly sheared center of the disk is unaffected, while material from the regions affected by FIC recrystallizes at a higher temperature, with material from the shish region showing distinctly faster recrystallization. Formation of shish in the rheometer was also detected before cooling by their effect on low-frequency viscoelastic response. Dynamic rheology of samples with sufficient shear applied above the threshold stress for shish formation show gel-like behavior at low frequency. Consistent with these findings, an abrupt increase in apparent viscosity was observed in flow through a capillary die at sufficiently high wall shear stress. FIC effects are most pronounced in materials with a small fraction of very long chains. In our samples we assign a “maximum molecular weight” Mmax based on the exponential decay of the high-molecular-weight tail measured in sizeexclusion chromatography. We find that samples with Mmax at or below 104 kg/mol show only weak or no FIC effects. We speculate that these very long chains may be necessary to form bridges between particulate impurities in the process of FIC precursor formation.

■

morphologies for iPP4, 5, and 6 (analogous to Figure 6 for iPP3), shear rate dependence of pressure for capillary rheometry of iPP6 (PDF)

Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Graham, R. S.; Olmsted, P. D. Coarse-grained simulations of flow-induced nucleation in semicrystalline polymers. Phys. Rev. Lett. 2009, 103, 115702. (2) Housmans, J.-W.; Steenbakkers, R. J.; Roozemond, P. C.; Peters, G. W.; Meijer, H. E. Saturation of point like nuclei and the transition to oriented structures in flow-induced crystallization of isotactic polypropylene. Macromolecules 2009, 42, 5728−5740. (3) Graham, R. S.; Olmsted, P. D. Kinetic Monte Carlo simulations of flow-induced nucleation in polymer melts. Faraday Discuss. 2010, 144, 71−92. (4) Hamad, F. G.; Colby, R. H.; Milner, S. T. Lifetime of FlowInduced Precursors in Isotactic Polypropylene. Macromolecules 2015, 48, 7286−7299. (5) Liu, D.; Tian, N.; Huang, N.; Cui, K.; Wang, Z.; Hu, T.; Yang, H.; Li, X.; Li, L. Extension-induced nucleation under near-equilibrium conditions: the mechanism on the transition from point nucleus to shish. Macromolecules 2014, 47, 6813−6823. (6) Wang, D.; Shao, C.; Zhao, B.; Bai, L.; Wang, X.; Yan, T.; Li, J.; Pan, G.; Li, L. Deformation-induced phase transitions of polyamide 12 at different temperatures: an in situ wide-angle X-ray scattering study. Macromolecules 2010, 43, 2406−2412. (7) Liu, Y.; Cui, K.; Tian, N.; Zhou, W.; Meng, L.; Li, L.; Ma, Z.; Wang, X. Stretch-Induced Crystal-Crystal Transition of Polybutene1: An in Situ Synchrotron Radiation Wide-Angle X-ray Scattering Study. Macromolecules 2012, 45, 2764−2772. (8) Hamad, F. G.; Colby, R. H.; Milner, S. T. Onset of FlowInduced Crystallization Kinetics of Highly Isotactic Polypropylene. Macromolecules 2015, 48, 3725−3738. (9) Hamad, F. G.; Colby, R. H.; Milner, S. T. Transition in Crystal Morphology for Flow-Induced Crystallization of Isotactic Polypropylene. Macromolecules 2016, 49, 5561−5575. (10) Kimata, S.; Sakurai, T.; Nozue, Y.; Kasahara, T.; Yamaguchi, N.; Karino, T.; Shibayama, M.; Kornfield, J. A. Molecular basis of the shish-kebab morphology in polymer crystallization. Science 2007, 316, 1014−1017. (11) Shen, B.; Liang, Y.; Kornfield, J. A.; Han, C. C. Mechanism for shish formation under shear flow: An interpretation from an in situ morphological study. Macromolecules 2013, 46, 1528−1542. (12) Ma, Z.; Balzano, L.; van Erp, T.; Portale, G.; Peters, G. W. Short-term flow induced crystallization in isotactic polypropylene: How short is short? Macromolecules 2013, 46, 9249−9258. (13) Hsiao, B. S.; Yang, L.; Somani, R. H.; Avila-Orta, C. A.; Zhu, L. Unexpected shish-kebab structure in a sheared polyethylene melt. Phys. Rev. Lett. 2005, 94, 117802. (14) Roozemond, P. C.; Ma, Z.; Cui, K.; Li, L.; Peters, G. W. Multimorphological crystallization of Shish-Kebab structures in isotactic polypropylene: quantitative modeling of parent-daughter crystallization kinetics. Macromolecules 2014, 47, 5152−5162. (15) Yang, H.-R.; Lei, J.; Li, L.; Fu, Q.; Li, Z.-M. Formation of interlinked shish-kebabs in injection-molded polyethylene under the

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.macromol.8b00563. Oscillatory shear cooling scans, POM crystal morphologies of iPP2 after various shear rates, crystal K

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX

Article

Macromolecules

criterion for orientation in polymer crystallization. Macromolecules 2008, 41, 1901−1904. (37) An, H.; Li, X.; Geng, Y.; Wang, Y.; Wang, X.; Li, L.; Li, Z.; Yang, C. Shear-induced conformational ordering, relaxation, and crystallization of isotactic polypropylene. J. Phys. Chem. B 2008, 112, 12256−12262. (38) Somani, R. H.; Yang, L.; Hsiao, B. S.; Fruitwala, H. Nature of Shear-Induced Primary Nuclei in iPP Melt. J. Macromol. Sci., Part B: Phys. 2003, 42, 515−531. (39) Pantani, R.; Balzano, L.; Peters, G. W. Flow-Induced Morphology of iPP Solidified in a Shear Device. Macromol. Mater. Eng. 2012, 297, 60−67. (40) Mykhaylyk, O. O.; Chambon, P.; Impradice, C.; Fairclough, P. A.; Terrill, N. J.; Ryan, A. J. Control of Structural Morphology in Shear-Induced Crystallization of Polymers. Macromolecules 2010, 43, 2389−2405. (41) Cogswell, F. N. In Polymer Melt RheologyA Guide to Industrial Practice; Han, C. D., Ed.; John Wiley & Sons: New York, 1981. (42) Hamad, F. G. Flow-Induced Crystallization in Isotactic Polypropylene. Doctoral Dissertation, Penn State University, 2015. (43) Cross, M. M. Rheology of non-Newtonian Fluids: A new flow equation for pseudoplastic systems. J. Colloid Sci. 1965, 20, 417−437. (44) Liu, Z.; Song, L.; Dai, X.; Yang, G.; Han, B.; Xu, J. Grafting of methyl methylacrylate onto isotactic polypropylene film using supercritical CO2 as a swelling agent. Polymer 2002, 43, 1183−1188. (45) Nazari, B.; Nazockdast, H.; Katbab, A. A. The role of flowinduced microstructure in rheological behavior and nonisothermal crystallization kinetics of polyethylene/organoclay nanocomposites. Polym. Eng. Sci. 2014, 54, 1839−1847. (46) Supaphol, P.; Spruiell, J. E. Nonisothermal bulk crystallization studies of high density polyethylene using light depolarizing microscopy. J. Polym. Sci., Part B: Polym. Phys. 1998, 36, 681−692. (47) Di Lorenzo, M. L.; Silvestre, C. Non-isothermal crystallization of polymers. Prog. Polym. Sci. 1999, 24, 917−950. (48) Zhang, Y. F.; Xin, Z. Isothermal and nonisothermal crystallization kinetics of isotactic polypropylene nucleated with substituted aromatic heterocyclic phosphate salts. J. Appl. Polym. Sci. 2006, 101, 3307−3316. (49) Lorenzo, A. T.; Müller, A. J. Estimation of the nucleation and crystal growth contributions to the overall crystallization energy barrier. J. Polym. Sci., Part B: Polym. Phys. 2008, 46, 1478−1487. (50) Xu, J. Z.; Chen, C.; Wang, Y.; Tang, H.; Li, Z. M.; Hsiao, B. S. Graphene Nanosheets and Shear Flow Induced Crystallization in Isotactic Polypropylene Nanocomposites. Macromolecules 2011, 44, 2808−2818. (51) Wang, J.; Yang, J.; Deng, L.; Fang, H.; Zhang, Y.; Wang, Z. More dominant shear flow effect assisted by added carbon nanotubes on crystallization kinetics of isotactic polypropylene in nanocomposites. ACS Appl. Mater. Interfaces 2015, 7, 1364−1375. (52) Wang, S. Q.; Ravindranath, S.; Wang, Y.; Boukany, P. New theoretical considerations in polymer rheology: Elastic breakdown of chain entanglement network. J. Chem. Phys. 2007, 127, 064903. (53) Boukany, P. E.; Wang, S. Q.; Wang, X. Step shear of entangled linear polymer melts: New experimental evidence for elastic yielding. Macromolecules 2009, 42, 6261−6269. (54) Dealy, J. M.; Wang, J. In Melt Rheology and Its Applications in Plastics Industry, 2nd ed.; Springer: Dordrecht, Netherlands, 2013. (55) Petraglia, G.; Coen, A. The Rheology of Polypropylene- The Influence of Molecular Weight Distribution on Melt Elasticity. Polym. Eng. Sci. 1970, 10, 79−85. (56) Minoshima, W.; White, J. L.; Spruiell, J. E. Experimental Investigation of the Influence of Molecular Weight Distribution on the Rheological Properties of Polypropylene Melts. Polym. Eng. Sci. 1980, 20, 1166−1176. (57) Ariff, Z. M.; Azlan, A.; Jikan, S. S.; Rahim, N. A. A. Rheological behaviour of polypropylene through extrusion and capillary rheometry. In Polypropylene; Dogan, F., Ed.; InTech: Rijeka, Croatia, 2012; Chapter 3, pp 29−48.

coexistence of lightly cross-linked chain network and oscillation shear flow. Macromolecules 2012, 45, 6600−6610. (16) Kanaya, T.; Matsuba, G.; Ogino, Y.; Nishida, K.; Shimizu, H. M.; Shinohara, T.; Oku, T.; Suzuki, J.; Otomo, T. Hierarchic structure of shish-kebab by neutron scattering in a wide Q range. Macromolecules 2007, 40, 3650−3654. (17) Balzano, L.; Ma, Z.; Cavallo, D.; van Erp, T. B.; FernandezBallester, L.; Peters, G. W. Molecular aspects of the formation of shish-kebab in isotactic polypropylene. Macromolecules 2016, 49, 3799−3809. (18) Kmetty, A.; Barany, T.; Karger-Kocsis, J. Self-reinforced polymeric materials: a review. Prog. Polym. Sci. 2010, 35, 1288−1310. (19) Zhong, G. J.; Li, L. B.; Mendes, E.; Byelov, D.; Fu, Q.; Li, Z. M. Suppression of Kin-Core Structure in Injection-Molded Polymer Parts by Situ Incorporation of a Microfibrillar Network. Macromolecules 2006, 39, 6771−6775. (20) Wang, Y.; Pan, J. L.; Mao, Y. M.; Li, Z. M.; Li, L. B.; Hsiao, B. S. Spatial Distribution of g-Crystals in Metallocene-Made Isotactic Polypropylene Crystallized Under Combined Thermal and Flow Fields. J. Phys. Chem. B 2010, 114, 6806−6816. (21) Roozemond, P. C.; van Drongelen, M.; Verbelen, L.; Van Puyvelde, P.; Peters, G. W. Flow-induced crystallization studied in the RheoDSC device: Quantifying the importance of edge effects. Rheol. Acta 2015, 54, 1−8. (22) Vleeshouwers, S.; Meijer, H. E. A rheological study of shear induced crystallization. Rheol. Acta 1996, 35, 391−399. (23) Nazari, B.; Rhoades, A. M.; Schaake, R. P.; Colby, R. H. FlowInduced Crystallization of PEEK: Isothermal Crystallization Kinetics and Lifetime of Flow-Induced Precursors during Isothermal Annealing. ACS Macro Lett. 2016, 5, 849−853. (24) Pogodina, N. V.; Winter, H. H.; Srinivas, S. Strain effects on physical gelation of crystallizing isotactic polypropylene. J. Polym. Sci., Part B: Polym. Phys. 1999, 37, 3512−3519. (25) Sadiku, E. R. Thermal studies on isotactic polypropylene. Measurements of Tg, Tm and ΔHf. Acta Polym. 1990, 41, 246−251. (26) Iijima, M.; Strobl, G. Isothermal Crystallization and Melting of Isotactic Polypropylene Analyzed by Time- and TemperatureDependent Small-Angle X-Ray Scattering Experiments. Macromolecules 2000, 33, 5204−5214. (27) Scelsi, L.; Mackley, M. R. Rheo-optic flow-induced crystallization of polypropylene and polyethylene within confined entry-exit flow geometries. Rheol. Acta 2008, 47, 895−908. (28) Farah, M.; Bretas, R. E. S. Characterization of i-PP ShearInduced Crystallization Layers Developed in a Slit Die. J. Appl. Polym. Sci. 2004, 91, 3528−3541. (29) Derakhshandeh, M.; Jazrawi, B.; Hatzikiriakos, G.; Doufas, A. K.; Hatzikiriakos, S. G. Flow-induced crystallization of polypropylenes in capillary flow. Rheol. Acta 2015, 54, 207−221. (30) Janeschitz-Kriegl, H.; Ratajski, E.; Stadlbauer, M. Flow as an effective promotor of nucleation in polymer melts: a quantitative evaluation. Rheol. Acta 2003, 42, 355−364. (31) Somani, R. H.; Yang, L.; Hsiao, B. S. Effects of high molecular weight species on shear induced orientation and crystallization of isotactic polypropylene. Polymer 2006, 47, 5657−5668. (32) Krishnaswamy, R. K.; Yang, Q.; Fernandez-Ballester, L.; Kornfield, J. A. Effect of the Distribution of Short-Chain Branches on Crystallization Kinetics and Mechanical Properties of HighDensity Polyethylene. Macromolecules 2008, 41, 1693−1704. (33) Hsiao, B. S. In Progress in Understanding of Polymer Crystallization; Reiter, G., Strobl, G. R., Eds.; Springer: New York, 2007; pp 133−149. (34) Janeschitz-Kriegl, H.; Ratajski, E. Some fundamental aspects of the kinetics of flow-induced crystallization of polymers. Colloid Polym. Sci. 2010, 288, 1525−1537. (35) Ratajski, E.; Janeschitz-Kriegl, H. Flow-induced crystallization in polymer melts: on the correlation between nucleation and specific work. Polym. Bull. 2012, 68, 1723−1730. (36) Mykhaylyk, O. O.; Chambon, P.; Graham, R. S.; Fairclough, J. P. A.; Olmsted, P. D.; Ryan, A. J. The specific work of flow as a L

DOI: 10.1021/acs.macromol.8b00563 Macromolecules XXXX, XXX, XXX−XXX