Dynamic viscoelastic results of 23 noncommercial metallocene-catalyzed polyethylenes and poly(ethylene/1-hexene) copolymers, in the range 130−190 °C, are presented. The effects of well-determined structural parameters such as molecular weight, polydispersity, and degree of short chain branching (SCB), are analyzed. The molecular weight varies between 60 000 and 325 000, the polydispersity between 1.8 and 7.3, and SCB between 0 and 48.5 branches/1000 C atoms. It is observed that a group of 11 polymers displays rheological specific features which can be summarized as follows:  (a) higher dynamic viscosities at low frequencies than other polyethylenes and ethylene/1-hexene copolymers of similar molecular weight, polydispersity and SCB degree; (b) higher relaxation times than narrow molecular weight distribution polyethylenes of similar dynamic viscosities at low frequencies but similar relaxation times to those of broad molecular weight distribution; (c) higher values of elastic modulus, in comparison with polyethylenes of similar molecular weight, polydispersity, and SCB but of the same order of magnitude as those of broader molecular weight distribution; (d) higher activation energy of flow than linear polyethylenes of the same molecular weight, polydispersity, and SCB level. An analysis of the literature results leads us to suspect that the polymers which show a “dissident” behavior possess a certain very low degree of long chain branching (LCB). The analysis of the samples by SEC coupled with intrinsic viscosimetry reveals that some of these 11 polymers are long chain branched. However, this technique does not appear to be enough sensitive to detect very small amounts of LCB, and an alternative single rheological method, based on the effect of temperature on dynamic viscosity, is proposed to evaluate the possible presence of LCB.

A nanocomposite sample was prepared by melt mixing a high density polyethylene (HDPE) with an in situ polymerized HDPE/multi wall carbon nanotube (MWNT) masterbatch. The nanocomposite had an approximate content of 0.52 wt % MWNT. Rheological, thermal, and mechanical properties were investigated for both neat HDPE and nanocomposite. The nanocomposite, when compared to the neat polymer, exhibits lower values of viscosity, shear modulus and shear stress in extrusion and a concurrent delay of the distortion regimes to higher shear stresses and rates. The nanocomposite presents also improved dimensional stability after processing, and lower values of the melt strength, draw ratio and viscosity in elongational flow. This behavior has been observed in composites in which an adsorption of a fraction (that with the highest molecular weight or relaxation time) of the polymer chains is considered. Furthermore, the enhancement in the crystallization kinetics, probed by rheometry and DSC, suggests that the carbon nanotubes act as nucleating agents for the polymeric chains. Additionally, the presence of adsorbed chains does not only influence the molten state but also induces interesting effects in the mechanical properties of the polymer. As a result, an increase of up to 100% in elastic modulus was observed in the HDPE/MWNT nanocomposite without losing the ductility present in neat HDPE.

Long molecular dynamics simulations of the melt dynamics, glass transition and nonisothermal crystallization of a C 192 polyethylene model have been carried out. In this model, the molecules are sufficiently long to form entanglements in the melt and folds in the crystalline state. On the other hand, the molecules are short enough to enable the use of atomistic simulations on a large scale of time. Two force fields, widely used for polyethylene, are taken into account comparing the simulation results with a broad set of literature experimental data. Although both force fields are able to capture the general physics of the system, TraPPe-UA is in a better quantitative agreement with the experimental data. According with the simulation results some fundamental aspects of polyethylene physical parameters are discussed such as the characteristic ratio (C n = 8.2 and 7.6 at 500 K, for TraPPe-UA and PYS force fields, respectively), the isothermal compressibility (α = 8.57 × 10 −4 K −1 ), the static structure factor and the melt dynamics regimes corresponding to an entangled polymer. Furthermore, the simulated T g (187.0 K) obtained for linear PE is in a very good agreement with the extrapolated T g values (185−195 K) using the Gordon−Taylor equation. Finally, the simulation of the nonisothermal crystallization process supports the view of a mixed state of adjacent and nonadjacent re-entry model. The simulated two phase model reproduces very well the initial fold length expected for high supercoolings and the segregation of the system in ordered and disordered layers. The paper highlights the importance of combining simulation techniques with experimental data as a powerful means to explain the polymer physics.

SynopsisThe melt rheology of ultrahigh molecular weight polymeric materials characterized by a narrow molecular weight distribution has been analyzed. Ultrahigh molecular weight polyethylene obtained from a metallocene catalyst shows a well-developed ''plateau'' modulus in a range of angular frequency of more than 3 decades. The characteristic value of the plateau modulus ͑ ϳ 2 MPa͒ is in close agreement with those reported for a model high molecular weight monodisperse polyethylene. From this value one can determine a characteristic molecular weight between entanglements of 1200 g mol Ϫ1. The molecular weight dependency of different, experimentally based relaxation times obtained from the linear viscoelastic response exhibits an exponent power law close to 3.0 for these materials. This seems to contradict the 3.4 dependence observed in the usual molecular weight range, which is based on the chain contour length fluctuation approach, but is in agreement with the latest reptation-based models. These models predict a crossover from the 3.4 to a 3.0 exponent for very long chains as used here at a constant critical value of the molecular weight M r close to 100M c (200M c when using the well accepted relationship M c ϭ 2M e ). This predicted crossover is independent of the polymer's chemical composition. However, combining results from our experiments with results from literature shows that the experimental values of M r extend from 15M c for polystyrene, 25M c for polyisobutilene, 100M c for polybutadyene to 220M c for polyethylene. These results are not predicted by molecular models and demand for new theoretical considerations of chain dynamics, in which the chemical structure is, most probably, a key factor that should be taken into account. It should be noticed that the influence of the molecular weight distribution on the differences observed is not understood. Unfortunately, it is very difficult to obtain monodisperse samples of ultrahigh molecular weight polyethylene and, therefore, the use of the samples studied here the best choice possible, up to now to test and revisit basic and novel aspects of the rheology of polyolefin's.

A combined computer simulation and experimental study describing the viscoelastic properties of linear polyethylene is presented. For the simulation, a set of C1000 polyethylene models were equilibrated using advanced Monte Carlo moves. Then, MD trajectories were calculated. From these simulations the entanglement molecular weight, M e, and the entanglement relaxation time, τe, were directly obtained. By introducing the experimental value for the plateau modulus and the simulated values for M e and τe into the reptation model, one finds that the derived curves of the relaxation shear modulus nicely coincide with the experimental ones.