Effect of Thermal History and Shear on the Viscoelastic Response of iPP Containing an Oxalamide-Based Organic Compound.

We report on the role of temperature and shear on the melt behavior of iPP in the presence of the organic compound N1,N1'-(propane-1,3-diyl)bis(N2-hexyloxalamide) (OXA3,6). It is demonstrated that OXA3,6 facilitates a viscosity suppression when it resides in the molten state. The viscosity suppression is attributed to the interaction of iPP chains/subchains with molten OXA3,6 nanoclusters. The exact molecular mechanism has not been identified; nevertheless, a tentative explanation is proposed. The observed viscosity suppression appears similar to that encountered in polymer melts filled with solid nanoparticles, with the difference that the OXA3,6 compound reported in this study facilitates the viscosity suppression in the molten state. Upon cooling, as crystal growth of OXA3,6 progresses, the decrease in viscosity is suppressed. Retrospectively, segmental absorption of iPP chains on the surface of micrometer-sized OXA3,6 crystallites favors the formation of dangling arms, yielding OXA3,6 crystallites decorated with partially absorbed iPP chains. In other words, the resulting OXA3,6 particle morphology resembles that of a hairy particle or a starlike polymer chain. Such hairy particles effectively facilitate a viscosity enhancement, similar to branched polymer chains. This hypothesis and its implications for the shear behavior of iPP are discussed and supported using plate-plate rheometry and slit-flow experiments combined with small-angle X-ray scattering analysis.

Introduction

Identification and optimization of the nucleation and crystal growth processes in semicrystalline polymers such as isotactic polypropylene (iPP) have been investigated over decades.[1] One route to enhance the nucleation process under quiescent conditions is the introduction of (foreign) organic or inorganic components that facilitate heterogeneous nucleation, i.e., the addition of a nucleating agent (NA).[2] Another route involves the application of shear flow which can generate pointlike nuclei that, in turn, can develop into row nuclei by alignment and aggregation until finally a shish is formed.[3,4] The generation of shear-induced precursors lowers the barrier for nucleation and, dependent on the applied shear and time, stimulates crystal growth either perpendicular to the shear direction or in an isotropic fashion.[5−7] The application of shear on polymer melts in the presence of nucleating fillers can enhance the nucleation process even further: Shear enhances the velocity gradient close to the filler, resulting in a local enhancement of the crystallization kinetics.[8] Additionally, in particular for nanoparticle filled systems, adsorption of polymer chains on the particle surface is found to influence the viscoelastic response and thus the shear behavior of the material.[9]

Hydrogen-bonding organic compounds, such as the aliphatic oxalamide N1,N1′-(propane-1,3-diyl)bis(N2-hexyloxalamide) (OXA3,6, Figure A), have gained quite some attention over the past years due to their ability to suppress the nucleation barrier for various polymers.[10−23] To be an efficient NA, these molecules are designed to (1) dissolve in a polymer matrix during processing and (2) crystallize upon cooling. The dissolution and crystallization process is found to be dependent on the OXA3,6 concentration, as shown in Figure B. Additionally, the nucleating efficiency of the polymer is found to be dependent on the NA particle size, size distribution, shape, and the time given for OXA3,6 crystallization: Combined, these parameters govern the surface-to-volume ratio of the dispersed nucleating agent. Because the self-assembly of OXA3,6 proceeds via a nucleation and growth mechanism, the resultant particle morphology is found to be strongly dependent on parameters such as the application of shear, concentration, and cooling rates (Figure C).[10,24]

Figure 1

(A) Chemical structure of oxalamide based nucleating agents. (B) Phase diagram showing the self-assembly of OXA 2,6, OXA 3,6, and OXA 4,6 in iPP during cooling (10 °C/min) as a function of NA concentration. (C) Polarization optical micrographs taken during isothermal crystallization of iPP in the presence of 0.5 wt % OXA3,6. Prior to isothermal crystallization, the sample was cooled from 200 to 170 °C at a rate of 30 °C/min (top) and 1 °C/min (bottom). After reaching 170 °C, the sample was cooled at a rate of 30 °C/min to 135 °C, after which the crystallization morphology was monitored over time. Because of the poor visibility of the NA particles, yellow dotted lines are placed over borders of the NA aggregates formed during cooling at a rate of 1 °C/min. Images B and C are adapted with permission from refs (10) and (24), respectively.

In addition to the nucleating efficiency, recently we reported that OXA3,6 facilitates a suppression in the apparent viscosity of the polypropylene matrix.[24] With the help of molecular modeling, we established that the viscosity suppression does not originate from differences in molecular weight, topology, or a solvent effect. Instead, we suggested that the suppression in apparent viscosity, in the presence of OXA3,6, is similar to that observed in polymeric systems filled with nanoparticles such as carbon nanotubes or graphene or silica.[25−29] High molecular weight iPP chains were considered to be physically absorbed on the dispersed nucleating agent surface, thereby effectively withdrawing them from the polymer melt.[30] Because of the effective contribution of the high molar mass component in the structure development, iPP was found to be less sensitive to shear in slit flow (i.e., lowered Weissenberg number[31]) in the presence of OXA3,6, resulting in the formation of significantly thinner shear layers compared to the pure iPP sample.[24] However, despite the suppression in apparent viscosity, systems containing OXA3,6 exhibited a decrease in the crystallization time and a more homogeneous crystallization morphology. Though these findings are interesting, a clear understanding on the origin and correlation between the viscosity suppression and shear is not available. To address these points, in this study we report on a series of experiments where we identify the effects of OXA3,6 concentration, thermal history, and shear on the apparent viscosity suppression of iPP. For the shear experiments, we make use of a slit-flow device equipped with diamond windows, which allows us to subject the polymer melt to high but well-defined shear rates and to follow the time-resolved structure development using X-ray scattering.

Experimental Section

Material Preparation and Blending Process

N1,N1′-(Propane-1,3-diyl)bis(N2-hexyloxalamide), henceforth abbreviated as OXA3,6, was synthesized in a two-step condensation reaction through the dropwise addition of a solution 1,3-diaminopropane (7.4 g, 0.1 mol) in THF (60 mL) to a 10-fold excess of diethyloxalate (146 g, 1 mol). The intermediate product, ethyl 2-[3-[(2-ethoxy-2-oxoacetyl)amino]propylamino]-2-oxoacetate (OXA3), was allowed to precipitate during the reaction at room temperature for 24 h, after which it was isolated through filtration. Next, the product was dissolved in chloroform, heated, and hot filtered. White OXA3 powder was obtained after the evaporation of the chloroform solvent from the filtrate. In the second reaction step, 5 g of OXA3 was dissolved in 200 mL of chloroform together with 3.7 g of hexylamine, and the mixture was left to stir under refluxing conditions for 48 h. The final product OXA3,6 was isolated through filtration, followed by washing with chloroform (2×) and diethyl ether (2×) and drying in vacuo at 80 °C overnight. The pure OXA3,6 displayed a melting temperature of 231 °C as observed in DSC analysis performed at a heating rate of 10 °C/min.[10] Similar to the crystallization behavior displayed in Figure B, the melting temperature of OXA3,6 crystallites is suppressed when blended in iPP. Melting is observed at a constant heating rate of 10 °C/min in polarization optical microscopy mounted with a Linkam hotstage at 185, 205, and 210 °C for blends having a concentration of 0.25, 0.5, and 1.0 wt % OXA3,6, respectively. The polypropylene used in this study was Borealis iPP (HD234CF grade). Prior to processing, all materials were dried overnight in vacuo at 80 °C. All samples (having various concentrations of OXA3,6) were extruded in a twin-screw extruder (DSM Xplore MC 5, 5 mL barrel) at 220 °C for 3 min at 100 rpm. The extrudate was directly injection molded into rheometer plates (2 mm thickness, 25 mm diameter) using a DSM Xplore, IM 5.5 microinjection molder. Furthermore, a second batch of polymer was processed following the same protocol, except that these were compression molded and cut into bars of 1.5 × 6 × 160 mm3 that fit into the slit-flow device.

Characterization Methods

The linear viscoelastic behavior (G′ and G″) of the samples was determined using a TA Instruments Discovery HR 1 rheometer mounted with a parallel plate geometry (25 mm). Samples were loaded at 180 or 220 °C and cooled to 150 °C at a rate of 5 °C/min using a temperature ramp (1 rad/s, 1% strain). Next, the sample was heated to 250 °C at a rate of 5 °C/min, whereas a frequency sweep was performed every 10 °C (1% strain). Note that frequency sweeps were only taken at 160 °C and higher as slow crystallization proceeded during frequency sweeps performed at 150 °C.

Polarization optical micrographs (between cross-polarizers and using a 530 nm λ-wave plate) were taken on an Olympus BX53 microscope mounted with an Olympus DP26 camera and a Linkam hotstage. The samples were heated to 220 °C at a rate of 30 °C/min and kept for 5 min under this condition. Next, the samples were cooled to 180 °C at a rate of 1, 5, or 30 °C/min to induce the crystallization of OXA3,6 into rhombic crystallites with various sizes. In some cases, to identify what happens to the OXA3,6 particles when subjected to shear stress, the sample was subjected to mild shear by hand after reaching 180 °C, and the optical morphology was monitored.

Online 2D small-angle X-ray scattering (SAXS) and wide-angle X-ray diffraction (WAXD) experiments (wavelength = 0.1 nm), following the morphological development during and after the flow, were performed in the European Synchrotron Radiation Facility (ESRF) at the DUBBLE beamline (BM26B, Grenoble, France). This beamline is particularly optimized for polymer science and allows for the installment of the slit-flow device, as is reported by Bras et al.[32] and Portale et al.[33] SAXS images were collected with a 2D Pilatus 1M detector (169 mm × 179 mm active area) placed at 3.50 m distance from the sample, whereas the WAXD images were collected using a Pilatus 300K detector (1472 × 195 pixels of 172 μm × 172 μm) at a distance of 0.21 m from the sample. The detector was triggered by an electric TTL pulse at the start of the displacement of the pistons to synchronize the data acquisition. The data were background corrected and normalized for synchrotron beam fluctuations using an ionization chamber placed before the sample. Furthermore, correction for the sample absorption was performed using a photodiode located at the beamstop. Both silver behenate and aluminum were used to calibrate the modulus of the scattering vector q-scale, using the position from diffraction peaks where q = 4π sin θ/λ with θ being half of the scattering angle.

Shear and Crystallization Protocols Using Slit-Flow Experiments

As is highlighted in Figure B, the phase behavior of OXA3,6 in iPP under quiescent conditions has already been reported. However, as mentioned in the Introduction, the nucleation efficiency of iPP in the presence of OXA3,6 and shear is significantly enhanced, although the exact origin is unknown. To accurately assess the effect of both shear and OXA3,6 on the viscoelastic response of iPP, slit-flow experiments were performed in the melt using a slit-flow device.[34] In this setup, the polymer is confined between two hydraulically driven pistons. Slit-flow experiments are performed by moving both pistons in the same direction. Similarly, pressurization of the polymers is imposed by moving the pistons toward each other, whereas the pressure is controlled up to 1000 bar and monitored by two pressure transducers. These pressure transducers are also used to monitor the pressure difference during slit-flow experiments. The structure development of the polymer upon slit-flow experiments is monitored online using X-ray diffraction through the two diamond windows having a 60° opening located in middle of the slit.

Through the application of shear pulse at different temperatures we assessed the effect of shear on the rheological behavior (Figure ) and structure development of iPP when OXA3,6 resides in the molten or crystalline state. Additionally, through the application of a series of pulses on the iPP melt, we can in turn assess how shear affects both the OXA3,6 particles and iPP shish formation. To probe these parameters, two slit-flow protocols were performed in this study: In the first slit-flow protocol we probe the effect of different thermal and shear histories on the behavior of the OXA3,6 crystallites and the resulting shish formation (Figure , left). In the second slit-flow protocol we probe the effect of the application of consecutive shear pulses at 160 °C on the shish formation and the development of the resulting shish-kebab morphology (Figure , right).

Figure 7

Pressure difference (ΔP) observed during the application of shear pulses at various temperatures for pure iPP (A), iPP + 0.5 wt % OXA3,6 (B), and 1.0 wt % OXA3,6 (C). Image (D) depicts an overview of the apparent viscosity during steady shear extracted from the data in images (A)–(C) using eqs –3.

Figure 2

Experimental protocols used in slit-flow experiments.

In more detail, slit-flow protocol 1 involves the application of a shear pulse at 150 °C (all pulses were 15 mm at 100 mm/s equaling an apparent wall shear rate γ̇wall of 400 s–1) after introducing three different thermal and flow histories: In the first step, the sample is loaded and heated to 180 °C to maintain the thermal history and OXA3,6 crystallite morphology induced during processing (i.e., OXA3,6 crystallites in samples having a concentration of 0.5 and 1.0 wt % OXA3,6 melt well above 180 °C). The sample is pressured to 50 bar, and a backward and forward shear pulse were given after a 3 min isotherm. The structure development was monitored over a period of 5 min. Next, the sample was cooled at a rate of 5 °C/min to 170 and 160 °C, where the same shear protocol was applied after reaching both temperatures. Lastly, in this step the sample was cooled to 150 °C after which a forward pulse was given and the shear morphology was monitored. Note that in this step of the protocol we do not melt the material in between shear pulses, indicating that we do not erase the flow history introduced at higher temperatures. The second step of the protocol involves the melting of the iPP crystallites formed at 150 °C without the melting of the OXA3,6 particles. This is done through heating to 180 °C, the application of a backward pulse to reset the slit-flow device, leaving the sample again for a 3 min isotherm. Now, the sample is immediately cooled to 150 °C, followed by a 3 min isotherm and a forward pulse. In the third step of this protocol, the sample is heated to 220 °C to erase all shear history and to melt the OXA3,6 particles. Upon reaching 220 °C, again a backward pulse is applied to reset the slit-flow device, followed by a 3 min isotherm. Next, the sample was cooled to 150 °C, a forward shear pulse (15 mm at 100 mm/s) was applied, and the structure development was monitored again for 5 min.

The second slit-flow protocol involves heating the sample to 180 °C to melt the iPP crystallites without melting the NA morphology introduced during processing. Next, the samples were cooled directly to 160 °C, after which a series of 10 shear pulses (5 forward and 5 backward pulses of 15 mm at 100 mm/s) were given with 1 min intervals. After the application of the 10 shear pulses, the sample was kept isothermal in order for iPP crystallization to proceed. Throughout this protocol, the structure development was monitored.

Results and Discussion

Viscosity Suppression of iPP in the Molten State of OXA3,6

The iPP grade used in this study (Borealis HD234CF) has an average molecular weight (Mw) of 310 kg/mol and a dispersity of Mw/Mn = 3.4 and was selected because its viscoelastic and shear responses have been extensively studied in the past.[35−38] To evaluate the effect of OXA3,6 on the linear viscoelastic response of this iPP grade, samples have been prepared with 0.25, 0.5, and 1.0 wt % OXA3,6. The samples were loaded in the rheometer at 220 °C to enforce melting of the OXA3,6 particles. A frequency sweep was performed after a 3 min isotherm under these conditions. Figures depicting the elastic modulus G′, the viscous modulus G″, and the phase angle of the frequency sweeps performed in this study are provided in the Supporting Information. From Figure it is apparent that the plateau in complex viscosity for the pure iPP sample is roughly 50% higher at 220 °C compared to the samples having OXA3,6. Additionally, it may be concluded that at 220 °C the complex viscosity decreases slightly with increasing OXA3,6 concentration throughout the evaluated frequency range. In line with previous observations,[24] these findings confirm that OXA3,6 facilitates a viscosity suppression in iPP.

Figure 3

Complex viscosity versus frequency for iPP containing various amounts of OXA3,6 determined at 220 °C. Note that two frequency sweeps are depicted per sample representing the high and low frequency range.

A comparable viscosity suppressing effect has been reported for polymers in the presence of solid nanofillers.[9] The difference in the viscosity suppression observed in this study is that it is facilitated while OXA3,6 is present in the molten state. Furthermore, the effect of OXA3,6 on the viscoelastic response of iPP is distinctly different from those observed in nanoparticle filled systems. As will be supported with slit-flow experiments, plate–plate rheometry, and polarized optical microscopy, OXA3,6 can facilitate both a suppression and enhancement of the viscoelastic response of iPP by a variation of the cooling rate while maintaining a constant OXA3,6 concentration. In contrast, the viscosity suppression in nanoparticle filled systems is only dependent on the concentration at a given particle size. What follows is a step by step explanation of the hypothesis, schematically represented in Figure , explaining both the viscosity suppressing and enhancement mechanism of iPP by OXA3,6.

Figure 4

Schematic representation of (left) micrometer sized OXA3,6 crystallites formed during slow cooling. Partial absorption of iPP chains leads to a decrease in chain mobility. Crystallites are localized in space, and the absorbed chains partially reorient via CLF. (middle) Partial absorption events leading to increase of chain mobility through molecular alignment. The OXA3,6 nanoclusters diffuse length scales of the average radius of gyration much faster than the iPP chains themselves do. (right) Formation of OXA3,6 crystallites of nano- and submicrometer size during rapid cooling. Nanocrystallites can contribute to the increase in chain mobility in a similar manner to nanoclusters. However, the larger crystallites have reduced mobility, effectively resulting in a decrease in the rate of absorption events responsible for the enhanced chain mobility. Furthermore, they might localize iPP chains, arresting their large scale reorientation. Note that the iPP chains in the bulk are not shown in this image.

With respect to the viscosity suppression, earlier, with the help of molecular modeling, we concluded that the OXA3,6 concentration is far too low to facilitate the observed enhancement in molecular relaxation, thus excluding the possibility that OXA3,6 acts as a solvent for iPP.[24] This implies that there should be some sort of interaction between the OXA3,6 molecules and the iPP chains, where the average effect is the speed-up of the chain dynamics, especially the large scale reorientation of the chains (i.e., in the terminal regime). According to the data provided in Figure , this interaction should not cause any significant disruption of the entanglement network. In polymer melts containing solid nanoparticles, the viscosity suppression has been interpreted as full withdrawal (absorption) of polymer chains on the particles.[9] Considering that iPP chains in this study are substantially entangled, full chain absorption would seem improbable in particular for the high molecular weight chains in the upper part of the molecular weight distribution.

Instead, we consider that OXA3,6 molecules do not fully dissolve in iPP upon melting but instead persist as nanosized clusters held together by weak hydrogen bonding persisting above the OXA3,6 melting temperature, an hypothesis in accordance with hydrogen bonding in aliphatic polyamides above their melting temperature.[39,40] In other words, following this hypothesis, when dissolved in iPP and above the OXA3,6 melting temperature, OXA3,6 persists as nanosized domains having only short-range order, rather than residing in a fully amorphous state. We speculate that the typical size of a nanocluster is somewhat smaller than the size of the entanglement mesh (tube diameter), meaning that it should contain only few (of the order of ten) OXA3,6 molecules. As a result, a nanocluster can diffuse distances of the order of the size of a typical iPP chain with molecular weight Mw (310 kg/mol) much faster than the iPP chain itself. In this respect, within the experimental time window, it can interact with numerous iPP segments of different chains. We hypothesize that the interaction suppresses the viscosity in the following manner: The affinity between the hexane end-groups of OXA3,6 and iPP segments as well as the loose arrangement of the clusters facilitates absorption/attachment of iPP segments (or small chain sections of size Me, the entanglement molecular weight). These absorbed segments temporarily carry some extra friction. Thus, molecular strands between absorbed segments can orient and stretch slightly. The segmental absorption is expected to be weak, meaning that it has a transient nature. Following the detachment of a nanocluster, the oriented strands will tend to rapidly rearrange themselves in a less aligned fashion. It should be noted that not all (partial) absorption events will produce the same degree of alignment. Furthermore, some events might produce alignment in directions other than the tube contour. In this respect, we consider that the observed viscosity suppression reflects the average effect of numerous absorption and detachment events occurring in the melt. Overall, the aforementioned interaction between dynamic OXA3,6 nanoclusters and iPP segments induces partial molecular alignment to the chains which, in turn, causes reduction of the chain friction and thus a speed-up of chain dynamics. Molecular alignment is known to reduce the chain friction at fast flows.[41] An alternative way to comprehend the proposed mechanism is to envisage a chain with an effective molecular weight that is slightly lower than the actual molecular weight. Figure , middle image, is an illustration of the proposed mechanism.

From the aforementioned theoretical standpoint, the fact that the viscosity suppression increases with decreasing frequency (as can be observed from Figure ) reflects the cumulative effect that the faster relaxation of shorter chains has on the relaxation of longer chains. That is, the OXA3,6 clusters speed up the relaxation of a chain of given Mw with respect to its relaxation in the pure iPP melt. In turn, this speed-up mediates a faster relaxation of chains of higher molecular weight through ordinary constraint release. The interaction of OXA3,6 nanoclusters with molecular segments of the latter chains enhances the chain’s dynamics over and above the former speed-up. Furthermore, with increasing dilution of the entanglement network, aligned strands of one chain might induce alignment on strands of neighboring chains.

Another possibility for the observed suppression in viscosity involves migration of OXA3,6 nanoclusters to the sample surface, facilitating slip with the rheometer plates during the experiments. However, the slit-flow experiments explained later in this work do not support this scenario.

Contribution of OXA3,6 Crystallites to the Viscoelastic Response of iPP

The observation of a viscosity enhancement of iPP by OXA3,6 requires the OXA3,6 to crystallize during cooling. Therefore, to investigate the influence of the crystallization of the OXA3,6 nanoclusters on the viscoelastic response, OXA3,6 nuclei were seeded by cooling the samples from 220 to 150 °C (Figure , left). Subsequently, the samples were heated to 160 °C and were subjected to a frequency sweep (Figure , right) under isothermal condition. As anticipated, a stepwise increase in the complex viscosity is observed on cooling the sample from 220 to 160 °C. The viscosity increase upon cooling of pure iPP solely reflects the temperature dependence of unperturbed reptation dynamics. However, in the presence of OXA3,6 the viscosity of the polymer melt exhibits a more abrupt upturn, which is related to the formation of OXA3,6 crystallites. From Figure (left) it is evident that the effective increase in complex viscosity upon crystallization of OXA3,6 is dependent on the OXA3,6 concentration; the higher the OXA3,6 concentration, the larger is the stepwise increase in complex viscosity. In addition, in accordance with Figure B, with increasing OXA3,6 concentration, the onset of crystallization of OXA3,6 particles proceeds at higher temperatures. For example, the OXA3,6 crystallizes at 180, 195, and 210 °C for iPP samples containing 0.25, 0.5, and 1.0 wt % OXA3,6, respectively.

Figure 5

(left) Temperature dependence of the complex viscosity for iPP containing various amounts of OXA3,6 during cooling from 220 to 150 °C at a rate of 5 °C/min, taken at a frequency of 1 rad/s and 1% strain. (right) Complex viscosity versus frequency for iPP containing various amounts of OXA3,6 determined at 160 °C. The black (pure iPP) and green (1.0 wt % OXA3,6) dotted lines show the predicted values of the complex viscosity obtained from the tube model, according to the scenario visualized in Figure S1A.

From the viewpoint of the partial absorption scenario presented above, the findings of Figure can be qualitatively explained as follows. For a given concentration, the OXA3,6 supersaturation, and thus the tendency to nucleate and crystallize, becomes larger with decreasing temperature. In other words, the ability of different nanoclusters to nucleate or aggregate into larger nanoclusters or crystallites increases. Thus, starting from the melting temperature, there is a temperature window where the average size of the formed OXA3,6 crystallites is still comparable to the size of the entanglement network; such nanoparticles can still contribute to the viscosity suppression through alignment of molecular strands. However, as crystal growth proceeds with decreasing temperature, their mobility drops as entanglements are influencing their diffusion. As a result, the overall frequency of partial absorption events decreases; such events begin to attain a local character, thereby influencing the configurations of fewer chains. Thus, compared to the pure melt, the overall effect might be a viscosity decrease or increase, depending on the average OXA3,6 crystal size (Figure , right). Upon further cooling, more and more nanoclusters will be taken up in the growing crystallites, which can reach micrometer dimensions if enough time is given. In this regime, the overall effect is a viscosity increase (relatively to the pure iPP melt) since a fraction of iPP melt chains has molecular segments absorbed on crystals that have dimensions much larger than the size of the chain (Figure , left). Moreover, the crystals are localized in space. Recall, however, that the weight fraction of OXA3.6 is very low; hence, we consider it unlikely that the viscosity increase, relative to the melt without OXA3.6, does originate from the crystals themselves. Instead, it arises from the slowdown of the molecular motions of the iPP chains that are partially absorbed on the crystals’ surface. With increasing concentration, the crossover to the regime of higher viscosity occurs at higher temperatures because crystals with dimensions beyond the nanoscale are formed already at higher temperatures.

Interestingly, for the sample having 0.25 wt % OXA3,6 the complex viscosity after OXA3,6 crystallization is still lower than that of the pure iPP. In contrast, the complex viscosity for samples having 0.5 and 1.0 wt % OXA3,6 surpasses the complex viscosity of the pure iPP throughout the evaluated frequency range (Figure , right). Moreover, at low frequencies a further increase in complex viscosity is observed, originating from the onset of a plateau in the elastic modulus (Figure S4). The origin of the plateau in elastic modulus can originate from (1) a bridging effect of iPP chains absorbed on the different OXA3,6 particles,[30] (2) the presence of long dangling iPP arms resulting from the physical absorption of iPP molecular segments on the OXA3,6 crystallites, or (3) a percolated network of OXA3,6 particles.[42]

The OXA3,6 rhombic crystallites grown under the conditions applied in the rheometer experiments are found to have a diagonal length of 10 μm (according to optical observations in between cross-polars presented in Figure S5). For comparison, the average radius of gyration of the polypropylene used in this study is 21.7 nm (estimated using Rg/Mw(1/2) = 0.039 nm).[43] Given this mismatch in length scales of the OXA3,6 crystallites and iPP coil sizes, we rule out the possibility of the iPP chains bridging the OXA3,6 crystallites.

The second possibility is in accordance with the partial absorption picture introduced above and schematically represented in Figure , left. When large rhombic crystallites are formed, the absorption of iPP molecular segments arrests reptative reorientation modes: Large-scale reorientation of the chains is frozen. Hence, the resulting OXA3,6 crystallites are expected to behave as starlike polymer chains or hairy particles in terms of their rheological response.[44−46]

In this scenario, the absorbed chains will contain long dangling arms. Such long dangling arms renew their orientation via contour length fluctuations (CLF) rather than reptation. For highly entangled chains, it is well established that the CLF relaxation mechanism is dramatically slower than reptation since the relaxation (reorientation) time of a chain segment along an arm has an exponential dependence on the arm molecular weight.[47,48] Such slowdown of the chain dynamics can give rise to the appearance of a broad shoulder or even a second plateau in the terminal regime.[49,50] Therefore, considering that chains in our systems have high molecular weights (recall Mw = 310 kg/mol, corresponding to 60–70 entanglements per chain), we anticipate a strong retardation in the terminal relaxation of the systems containing large, micrometer-sized NA crystallites. To support this hypothesis further, we used molecular modeling[47] to examine how the viscoelastic response changes in the presence of long dangling arms for the melt having 1.0 wt % OXA3,6. The system having 1.0 wt % OXA3,6 was chosen due to its larger NA concentration and should thus lead to a higher degree of absorption of chains having long dangling arms. As readily seen in Figure , right (green dashed line), the presence of a low volume fraction (i.e., 10%) of long dangling arms indeed changes the linear viscoelastic response significantly. At low frequencies, an upturn in the complex viscosity is predicted in agreement with the experimental data (green symbols). The agreement is not quantitative because the exact volume fractions and molecular weight distribution of the (bulk) linear chains and the dangling ends of the absorbed chains are unknown: The value of 10% was used as an estimate since the exact volume fraction of dangling arms is unknown. The same applies to the molecular weight distribution of the arms; thus, the latter was constructed from the high molecular weight tail of the pure iPP melt (see the Supporting Information for more details). Concerning the pure iPP melt, the tube model matches the experimental data (black solid line, Figure , right) by considering reptative dynamics. Note that the parametrization used to obtain the theoretical predictions in Figure is consistent with the one used in our earlier work.[24] Overall, the predictions of the tube model support the possibility that long dangling iPP arms on OXA3,6 crystallites are responsible for the rise in the low-frequency plateau in elastic modulus in the presence of OXA3,6 crystallites.

When the origin of the elastic plateau in the terminal regime is the result of such long iPP arms absorbed on OXA3,6 particles, we can expect that this behavior is triggered at a critical OXA3,6 particle size. As explained earlier, this critical particle size appears to be of order of the radius of gyration of the high molecular weight iPP chains. Apart from the OXA3,6 concentration, the net contribution of the OXA3,6 crystallites will also be dependent on the thermal history. These two factors together govern the crystallite nucleation and growth process. With respect to the thermal history, on slow cooling, the initial nucleation rate of OXA3,6 crystallites will be low due to the low undercooling. As is visible from Figure C, such low nucleation density results in the growth of large OXA3,6 crystallites. The number of nuclei will increase during further cooling, eventually resulting in a distribution of OXA3,6 crystallites with varying but large size. Indeed, as observed from Figure , right, and Figure S5, systems with large OXA3,6 crystallite size facilitate a strong increase in viscosity compared to pure iPP. Similarly, subjecting the sample to high cooling rates increases the nucleation density of OXA3,6 due to the rapid achievement of high undercooling, yielding a large number of small OXA3,6 crystallites (Figure C) with a net viscosity decrease compared to the iPP, as reported in earlier work.[24]

With respect to the used OXA3,6 concentration, as is visible from Figure B, increasing the OXA3,6 concentration increases the temperature where crystallization proceeds, thereby decreasing the nucleation density (lower supersaturation) and thus favoring the growth of large OXA3,6 crystallites.[10] Indeed, this feature is detected in our samples: The usage of 0.25 wt % OXA3,6 apparently yields sufficiently small crystallites to facilitate a net viscosity suppression compared to the pure iPP (Figure , right). In contrast, increasing the OXA3,6 concentration to 0.5 or 1.0 wt % increases the crystallite size such that it favors the generation of dangling iPP arms and the concomitant viscosity enhancement.

To assess the third possibility for the elastic network in the terminal regime (i.e., the presence of a percolated network of OXA3,6 crystallites), we examine whether a reduction of particle size leads to an increase in elasticity. As has been reported by Hao and co-workers,[51] the critical concentration for the generation of a particle network is achieved at lower concentrations when using particles with smaller size. Consequently, decreasing the particle size in an already percolated network should increase the elastic modulus of the particle network. To relate this finding to the current scope, recall that the samples were prepared in an extruder at 220 °C, after which they were injection molded into rheo-discs. The high shear and cooling rates applied during injection molding result in OXA3,6 particles with submicrometer dimensions as they are barely detectable in polarization optical microscopy (Figure S5). In practice, the samples having 0.5 and 1.0 wt % OXA3,6 could be loaded in the rheometer at 180 °C without melting the OXA3,6 particle morphology generated during processing. Therefore, loading the samples at 180° allows us to probe the effect of OXA3,6 crystallites with a significantly decreased size compared to when they are crystallized during cooling after loading at 220 °C. Once loaded in the rheometer, the materials were cooled to 150 °C, after which they were subjected to the same protocol as the samples loaded at 220 °C (i.e., heated to 160 °C, followed by a frequency sweep).

As is shown in Figure left, loading the sample having 0.5 wt % OXA3,6 at 180 °C, i.e., maintaining the fine OXA3,6 particle morphology generated during processing, yields in fact a lowered viscoelastic response compared to the pure iPP. The fact that a decrease in particle size results in a significant suppression in viscoelastic response excludes the possibility for the presence of an OXA3,6 particle network under the evaluated conditions. Instead, these findings favor the hypothesis that physical absorption of iPP segments on the OXA3,6 crystallite surface determines the viscoelastic response of the melt for two reasons: (1) the viscosity of the sample is lower than the pure iPP, suggesting that the OXA3,6 crystallites are sufficiently small to produce an overall reduction of chain friction via the partial absorption mechanism explained above, and (2) the presence of the onset of a plateau in the elastic modulus at low frequencies (Figure S6) suggests that these OXA3,6 crystallites also contain dangling iPP arms (hence, this example follows the absorption scenario depicted in Figure , right). Similar results are obtained for the sample having 1.0 wt % OXA3,6, for which the data are provided in Figure S7.

Figure 6

Effect of thermal history on the viscoelastic behavior for iPP containing 0.5 wt % OXA3,6 (left) and the temperature dependency of the complex viscosity as a function of thermal history, taken at a frequency of 1 rad/s and 1% strain (right). Note that the samples with OXA3,6 were loaded at 180 and 220 °C, whereas the iPP sample was loaded at 220 °C. All samples were cooled to 150 °C at a rate of 5 °C/min, followed by heating to 160 °C and a frequency sweep.

Although loading the sample at 180 °C results in a lowered complex viscosity compared to the pure iPP at 160 °C, the complex viscosity drops further upon melting of the OXA3,6 particles. As is shown in Figure , right, the complex viscosity of samples loaded at the different temperatures, i.e., samples having a different OXA3,6 particle size, coincides upon melting of the OXA3,6 particles and upon further heating. These findings indicate that the viscosity enhancing effect of the dangling iPP arms is lost upon melting of the OXA3,6 crystallites, irrespective of the thermal history of the sample. Again, similar behavior was observed for the sample containing 1.0 wt % OXA3,6, as is shown in Figure S8.

Confirming the Viscosity Suppression in iPP with OXA3,6 Using Slit-Flow Experiments

Slit-flow experiments were performed to identify whether the viscosity suppression in iPP with OXA3,6 persists during shear experiments. Therefore, samples having 0.5 and 1.0 wt % OXA3,6 were heated to the desired temperature, and a shear pulse of 15 mm was given at a rate of 100 mm/s to the samples. During the pulse, the pressure difference (ΔP in bar) was monitored, as can be seen from Figures A–C. Generally, an initial pressure buildup is observed at the start of the shear pulse until a steady shear plateau is achieved. The pressure difference during the pulse was converted to the apparent viscosity (η) using eqs –3:where W and H are the slit width and thickness of 6 and 1.5 mm, respectively, Q is the volumetric flow rate being H × W × Vpiston, and L is the distance between the pressure transducers (120 mm).[52] The piston speed (Vpiston) for shear pulses in this study was fixed to 100 mm/s, corresponding to an apparent shear rate (γ̇) of 400 s–1 (see eq ). The plateau viscosity during steady shear was plotted as a function of shear temperature and is shown in Figure D. Indeed, in line with previous observations, the OXA3,6 facilitates a viscosity suppression when it resides in the molten state. Furthermore, at temperatures below 180 °C, the steady shear viscosity of both samples having 0.5 and 1.0 wt % OXA3,6 start to increase and approach the viscosity of the pure iPP sample, which is attributed to crystallization of the OXA3,6 particles.

Effect of Shear on the OXA3,6 Morphology and Viscoelastic Behavior of iPP

Although the viscosity suppression in iPP with OXA3,6 is evident from shear experiments, we have observed that the OXA3,6 particles are brittle and highly sensitive to folding and breakup during shear (Figure ). Therefore, to establish a relation between shear, the viscosity suppressing effect, and OXA3,6 particle breakup, we have subjected the samples to slit-flow protocol 1 described in the Experimental Section. To recall, here we evaluate the effect of shear and thermal history using three different steps: In the first step we load the material at 180 °C and introduce thermal history to the sample through the application of consecutive shear pulses during cooling (i.e., pulses are given at 180, 170, 160, and 150 °C). In the second step, the sample is heated to 180 °C to erase the previously introduced thermal history while preventing the melting of the generated OXA3,6 particle morphology. After a 3 min isotherm, the sample is cooled to 150 °C, after which another shear pulse is given. The third and last steps of this protocol involve the removal of the thermal history of the sample and melting of the OXA3,6 crystallites through heating to 220 °C. Again, the samples are cooled to 150 °C, after which the last shear pulse is given.

Figure 8

(A) The image on the left depicts the OXA3,6 particle distribution as observed in polarization optical microscopy after cooling iPP containing 1.0 wt % OXA3,6 from 220 to 180 °C at a rate of 1 °C/min, whereas the image on the right depicts the folding and breakup of the OXA3,6 particles after drag of the top glass plate at 180 °C. To ease the detection of the characteristic OXA3,6 particle morphology before and after shear, color-based segmentation has been performed on images (A), whereas the results are depicted in (B).

Figure shows the 2D-SAXS signals depicting the equatorial scattering region obtained directly after the consecutive shearing steps performed in the first step of slit-flow protocol 1. In general, the application of these shear pulses does not result in the generation of any notable SAXS signal for the pure iPP. In contrast, for both the samples containing 0.5 and 1.0 wt % OXA3,6, a streak in the equatorial scattering intensity appears after applying a shear pulse at 170 °C and develops further upon application of the following shear pulses. Such a streak in equatorial scattering intensity suggests the presence of nanosized entities aligned along the shear direction, possibility originating from shear-induced iPP shishes or from the OXA3,6 particles themselves. Further cooling to 150 °C and the application of shear yields the characteristic shish-kebab morphology for all three samples, where the presence of kebabs is deduced from the lobes in vertical scattering intensity. Interestingly, both signals corresponding to the shishes and kebabs are significantly weaker in the pure iPP sample than for the samples containing OXA3,6. These findings indicate that despite the lowered apparent viscosity, iPP in the presence of OXA3,6 is more sensitive to undergo shear-induced crystallization.[53]

Figure 9

SAXS patterns depicting the equatorial region obtained during the shear pulses at 180, 170, and 160 °C applied in slit-flow protocol 1. Similarly, the SAXS patterns obtained after shearing at 150 °C during the various steps in slit-flow protocol 1 are provided. Note that the images taken directly after the shear pulse applied at 150 °C in step 1 display the full 2D SAXS patterns to visualize the vertical lobes in scattering intensity resulting from the presence of crystal kebabs.

More information about this shear behavior is obtained from the remainder of the slit-flow protocol: No kebabs are formed during the shear pulse at 150 °C in step 2 of slit-flow protocol 1. This is likely the result from the decreased number of shear pulses applied during cooling in step 2 of in this protocol. More importantly, in contrast to the pure iPP sample, heating to 180 °C does not remove the scattering in the equatorial region for the samples containing OXA3,6, which explains why it persists after cooling back to 150 °C. Only upon heating beyond 180 °C, a gradual decrease in the equatorial scattering signal is observed in the samples containing OXA3,6, until the signal fully disappears above 210 °C, as shown in Figure S9. Although it is reported in the literature that a small fraction of crystallites in shishes can persist above the polymers’ nominal melting temperature,[54,55] the equatorial scattering signal from the shear-induced shishes in the pure iPP is readily lost when reaching 180 °C. Accordingly, it is not likely that the streak in equatorial scattering intensity in samples containing OXA3,6 originates from regular shear-induced shishes. Given that melting of OXA3,6 crystallites also proceeds between 180 and 210 °C as shown in Figure and Figure S8, we consider it more likely that the equatorial scattering signal is related to the OXA3,6 particles themselves. Furthermore, after cooling from 220 °C and performing the shear pulse in step 3 of slit-flow protocol 1, we only observe a small equatorial streak in the samples containing OXA3,6, comparable to that of the shear-induced shishes formed in the pure iPP. This indicates that formation of the intense streak in equatorial scattering intensity in the presence of OXA3,6 requires multiple shear pulses.

A possible explanation for the observed shear behavior can be found in the combined shear alignment and particle breakup of the OXA3,6 crystallites during shear. As discussed in the previous section, OXA3,6 is expected to physically absorb iPP segments on its surface. OXA3,6 particles decorated with iPP chains are likely to cause a disturbance in the flow field and thereby stimulate shear-enhanced shish formation. In particular, when the OXA3,6 crystallites break up during shear, it is possible that the absorbed iPP chains remain anchored on different OXA3,6 fragments. Given that the OXA3,6 fragments for during shear, it is likely that the iPP chains anchored on different OXA3,6 fragments are stretched in the process, effectively facilitating a local bridging effect. Such bridging chains would be excellent precursors for shish growth; hence, OXA3,6 particle breakup is suggested to facilitate shear-enhanced shish formation. Furthermore, as the bridging chains are anchored to the OXA3,6 particles, bundles of bridging chains will persist well above the nominal melting temperature of iPP and give rise to the previously observed equatorial scattering signal in SAXS. This signal would only disappear when these stretched chains can relax, i.e., upon the melting of the OXA3,6 particles, as is in line with our SAXS observations. An example of the proposed mechanism is shown in Figure .

Figure 10

Schematic representation of the shear-induced OXA3,6 (red rhombic crystallites) fragmentation, resulting in shear-enhanced shish (black bundles) formation through stretching of anchored chains in addition to regular shear-induced shish formation. Additionally, given the nucleating ability of OXA3,6 particles, they facilitate kebab growth (blue lines) even further.[10]

To verify this hypothesis, a second slit-flow protocol was performed to obtain more information about how OXA3,6 affects the iPP behavior when subjected to multiple shear pulses. As explained in the Experimental Section (slit-flow protocol 2), samples were loaded at 180 °C, cooled, to 160 °C and subjected to 10 shear pulses with a 1 min interval (5 forward pulses, 5 backward pulses). Figure left shows the ΔP evolution during the shear pulses performed for the sample containing 1.0 wt % OXA3,6. The pulses for the pure iPP and the sample containing 0.5 wt % OXA3,6 are provided in Figure S10. Next, the ΔP during steady shear was converted to the apparent viscosity using eq . To highlight the difference in shear behavior, the apparent viscosity of the first iPP pulses were subtracted as normalization procedure: The apparent viscosity of the first iPP forward pulse was subtracted from the apparent viscosity of all forward pulses, whereas the apparent viscosity of the first iPP backward pulse was subtracted from the apparent viscosity of all backward pulses. Figure (right) shows the resulting overview.

Figure 11

Left: pressure difference (ΔP) observed during the application of the 10 consecutive shear pusles in slit-flow protocol 2, for the sample having 1.0 wt % OXA3,6. Right: the apparent viscosity difference calculated from the plateau in ΔP during steady shear using eq . Note that only the differences in the presented apparent viscosity values are presented as they are normalized through subtraction of the apparent viscosity of the first forward or backward pulses from the pure iPP sample.

For the pure iPP sample it can be observed that the steady shear viscosity slowly increases upon the application of the consecutive shear pulses. As is reported in the literature, shear can result in the generation of pointlike nuclei that together with stretched chain segments assemble and grow into a shish when sufficient shear is applied.[3,4] During the application of a shear pulse in the slit flow, the polymer at the wall will be subjected to the highest shear rate and will therefore generate shishes the fastest. As is suggested by Peters and co-workers, these shishes contain polymer chains that are partially embedded in the shish and partially located in the melt; i.e., the shishes are covered in “hairs” that interact with the surrounding melt.[50] During the application of a shear pulse, shishes are generated in sufficient concentration and with sufficient length to macroscopically decrease the fluidity of the sheared layer, thereby decreasing its deformation rate. As a result, every time a pulse is given in this slit-flow protocol, a small shear layer is deposited at the wall of the sample, effectively decreasing the remaining area for volumetric flow, thus increasing the pressure drop and the viscosity calculated therefrom. Such layer deposition at the wall with application of consecutive shear pulses explains the rather exponential increase in steady shear viscosity observed in the slit-flow protocol 2.

In line with previous observations, we observe that the samples containing OXA3,6 start with decreased viscosity compared to the pure iPP (Figure , right). However, upon the application of consecutive shear pulses the ΔP, and thus the viscosity, increases rapidly and surpasses that of the pure iPP. The sample containing 0.5 wt % OXA3,6 exhibits the same steady shear viscosity as the pure iPP after six shear pulses. Similarly, the sample containing 1.0 wt % OXA3,6 exhibits the same steady shear viscosity as the pure iPP already after four shear pulses. This indicates that application of shear in the presence of OXA3,6 generates thicker shear layers for each applied shear pulse. Indeed, when looking at the 2D-SAXS patterns provided in Figure , we can observe similar features as reported in Figure for samples containing OXA3,6; the consecutive application of shear pulses results in a rapid development in the streak in equatorial scattering intensity. Furthermore, in line with higher steady shear viscosity and thus the presence of the thicker shear layers, the samples containing OXA3,6 facilitate a more dominant kebab growth, as is also evident from the 2D-WAXD patterns provided in Figure S11. Additionally, these findings suggest that the suppression in iPP viscosity in the presence of OXA3,6 is facilitated throughout the whole sample: Recall, earlier we mentioned the possibility that the observed viscosity suppression is a result from sample slip originating from OXA3,6 migrated to the sample surface. Following this scenario, the deposition and immobilization of the first shear-layer in slit-flow experiments should negate this effect as the shear layer is considered immobilized during the following shear pulses. However, this is not the case in our samples; the viscosity suppression in the presence of OXA3,6 persists during the first 3–5 pulses, indicating that the remaining melt exhibits a decreased viscosity, in particular since the presence of OXA3,6 results in the generation of thicker shear layers than the pure iPP.

Figure 12

2D-SAXS patterns obtained directly after the application of shear pulses 1, 3, 5, and 10 applied in slit-flow protocol 2. Note that the scaling of the scattering intensity is the same as used for Figure .

Overall, these results are in line with our previous hypothesis: Application of shear in the presence of OXA3,6 results in significant enhancement in shish formation, followed by the enhanced development of the shish-kebab morphology. Given that the OXA3,6 particles break up during shear and are presumed to be capable of withdrawing iPP segments from the melt through physical absorption, the generation of stretched iPP chains between the fragmented OXA3,6 particles is a plausible origin for this shear-enhanced shish formation. This hypothesis is further supported by the fact that the signal in equatorial scattering intensity, originating from these iPP bridges, persists up to the melting temperature of the OXA3,6 crystallites. It is noteworthy that the shear-enhanced shish formation and the viscosity suppression in the presence of OXA3,6 have the same origin: They both originate from physical absorption of iPP segments on the OXA3,6 particles.

Conclusions

In this work we evaluated the influence of thermal history and shear on the melt behavior of iPP in the presence of the organic compound OXA3,6. We demonstrate that a viscosity suppression is observed in iPP when OXA3,6 resides in the molten state, which we attribute to an effective reduction of chain friction resulting from molecular alignment of iPP segments. We argued that the alignment is induced by partial absorption of iPP molecular segments on OXA3,6 nanoclusters present in the melt. However, upon cooling, the OXA3,6 nanoclusters develop into crystallites having dimensions beyond the radius of gyration of the iPP chains and limiting their ability to cause significant reduction of chain friction. Instead, the physical absorption of iPP segments on OXA3,6 crystallites favors the formation of dangling arms, effectively resulting in a viscosity enhancement compared to pure iPP and the generation of a plateau in elastic modulus. This effect is amplified when using increasing OXA3,6 concentrations or slow cooling rates as this increases the chance for the OXA3,6 nanoclusters to be taken up in the crystallites. In addition, the application of shear in the presence of OXA3,6 (1) enhances the shear layer formation, (2) fragments OXA3,6 crystallites, and (3) gives rise to an equatorial streak in scattering intensity that persists up to the melting of the OXA3,6 crystallites. This scattering signal is thought to result from a local bridging effect of physically absorbed iPP chains that are stretched between the fragmented OXA3,6 crystallites, resulting in shear-enhanced shish formation and explaining the enhanced shear response.