A Conditional Monte Carlo Method to Determine the Architectures of Metallocene Catalyzed Polyethylene

Two polymer molecules of the same length (n) and the same number of branch points (N) can have different properties, since they may possess distinct architectures. In this paper we present a conditional Monte Carlo algorithm for the virtual synthesis of metallocene‐catalyzed polyethylene (PE) in a continuous stirred tank reactor (CSTR). The condition for the Monte Carlo method consists of a fixed chain length distribution (CLD) and a degree of branching distribution (DBD). These distributions are calculated with a Galerkin finite element method. The synthesis method is a recursive algorithm that subsequently creates insertions of sub‐structures containing numbers of branch points according to a certain probability density function. This provides an adjacency matrix describing the connectivity between the branch points, while separately a vector containing the length of segments between branch points and terminal segments is generated. Characterization of the architectures proceeds by rheological features, seniorities and priorities, and molecular properties like the radius of gyration. Comparing the radii of gyration of metallocene polyethylene and low density PE (ldPE) shows the former to possess a more comb‐like structure on average. This is confirmed by the rheological characterization. The found bivariate seniority/priority distribution agrees well to the results of an analytical study of the same chemical system.

Constituting elements of the algorithm.     

Polyethylene Made with Combinations of Single‐Site‐Type Catalysts: Monte Carlo Simulation of Long‐Chain Branch Formation

The formation of long‐chain branches (LCBs) during ethylene polymerization with a combination of catalysts was studied by Monte Carlo simulation. The model describes polymerization with a non‐branching catalyst that produces linear macromonomers, and a branching catalyst that produces linear and branched macromonomers. The LCBs are formed when the branching catalyst incorporates a macromonomer. The discussion is based on the three types of chain topology obtained during the synthesis: linear, comb‐branched, or hyperbranched. Simulation results show how the chain length distribution and the number of LCBs change according to the ratio between the two catalysts present in the reactor. The ratio hyperbranched/comb‐branched is defined to evaluate the system composition and the contribution of each catalyst.     

Free‐Radical Polymerization with Long‐Chain Branching and Scission in a Continuous Stirred‐Tank Reactor

Free‐radical polymerization that involves the polymer transfer reactions leading to both long‐chain branching and scission, as in the cases of high‐pressure olefin polymerization, is considered. In CSTR, the residence time distribution is broad and the primary polymer chain, whose residence time is large, is subjected to polymer transfer reaction for a longer time, leading to a larger number of branching and scission points. The distributions of both branching and scission density are much broader in a CSTR than in a batch, or equivalently, a PFR. The radius of gyration for larger sized polymers formed in a CSTR tends to be much smaller than that for randomly branched polymers.

     

Shear damping function measurements for branched polymers

We present results for step‐strain experiments and the resulting damping functions of polyethylene blends of different structures, including solutions of linear, star and comb polymers. Remarkably, an entangled melt of combs exhibits a damping function close to that for entangled linear chains. Diluting the combs with faster‐relaxing material leads to a more nearly constant damping function. We find similar behavior for blends of commercial low density polyethylene LDPE. Our results suggest a simple picture: on timescales relevant to typical damping‐function experiments, the rheologically active portions of our PE combs as well as commercial LDPE are essentially chain backbones. When strongly entangled, these exhibit the Doi‐Edwards damping function; when diluted, the damping function tends toward the result for unentangled chains described by the Rouse model – namely, no damping. © 2007 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 45: 3117–3136, 2007     

Dimensions of branched polymers formed in simultaneous long‐chain branching and random scission

In free‐radical olefin polymerizations, the polymer‐transfer reactions could lead to chain scission as well as the formation of long‐chain branches. The Monte Carlo simulation for free‐radical polymerization that involves simultaneous long‐chain branching and random scission is used to investigate detailed branched structure. The relationship between the mean‐square radius of gyration 〈s²〉 and degree of polymerization P as well as that between the branching density and P is the same for both with and without random scission reactions—at least for smaller frequencies of scission reactions. The 〈s²〉 values were larger than those calculated from the Zimm–Stockmayer (Z‐S) equation in which random distribution of branch points is assumed, and therefore, the Z‐S equation may not be applied for low‐density polyethylenes. The elution curves of size exclusion chromatography were also simulated. The molecular weight distribution (MWD) calibrated relative to standard linear polymers is much narrower than the true MWD, and high molecular weight tails are clearly underestimated. A simplified method to estimate the true MWD from the calibrated MWD data is proposed. The MWD obtained with a light scattering photometer in which the absolute weight‐average molecular weight of polymers at each retention volume is determined directly is considered a reasonable estimate of the true MWD. © 2001 John Wiley & Sons, Inc. J Polym Sci Part B: Polym Phys 39: 2960–2968, 2001     

Computing the Trivariate Chain Length/Degree of Branching/Number of Combination Points Distribution for Radical Polymerization with Transfer to Polymer and Recombination Termination

Summary: Polymer molecules made by radical polymerization with transfer to polymer and recombination termination contain branch points (connecting branch arms to backbones) and combination points (connecting various molecular structures). The trivariate chain length/degree of branching/number of combination points distribution (CLD/DBD/CPD) was calculated using a two‐dimensional version of a previously used pseudo‐distribution approach. This yielded the CPD moments for given chain length n and number of branches i. Both DBD and CPD at given chain length resemble a binomial distribution. For the construction of the full DBD and CPD a set of orthogonal polynomials, the Krawtchouk polynomials, in combination with the binomial distribution was employed. A first‐order Krawtchouk approximation enabled to compute the full CLD/DBD/CPD from the three CPD moments as a function of n and i. Results agree well with those from a Monte Carlo (MC) simulation method. However, the large scatter due to the small numbers of molecules collected in the MC method at longer chain lengths prevents comparison in this range.

Solutions of 3D CLD/DBD/CPD: CPD at constant chain length (20 000) and number of branch points (20).     

Simulating Light Scattering Behavior of Branched Molecules

Monte Carlo simulations reveal long chain branching (LCB) topology based on kinetics of systems like low‐density polyethylene (ldPE). Examining the topologies computed shows the majority of branch arms to be short as compared to backbone length, while also a significant branch‐on‐branch fine structure is observed. Until now, predicting scattering function P ⁻¹(θ) from LCB has only been successful for structures like simple combs or stars. Topologies in graph theoretical form are used to predict scattering function P ⁻¹(θ) and by summating intramolecular distances, accounting for branching and excluded volume. Experimental size exclusion chromatography with multiangle light scattering can be brought in line with the predicted branching character. Branching is less than predicted due to a different fine structure in ldPE leading to stronger size contraction.     

Bimodal molecular weight distribution formed in the emulsion polymerization of ethylene

It is known that the molecular weight distribution (MWD) formed in an emulsion polymerization of ethylene can be bimodal. However, the origin of the bimodality has not been elucidated. In this article, a Monte Carlo simulation is conducted, mostly with parameters reported in the literature. The simulated MWDs are bimodal because of the limited volume effect; that is, the high molecular weight profiles are distorted by the small particle size, which is comparable to the size of the largest branched polymer molecule in a particle. The simulated MWDs agree reasonably well with the experimentally obtained MWDs. © 2002 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 40: 3426–3433, 2002     

Chain confinement in polymer nanocomposites and its effect on polymer bulk properties

The issue of chain confinement in nanocomposites remains largely unanswered because experimental systems are plagued by additional complicating variables such as particle–polymer interactions and free volume increases brought upon by the addition of the particles. Using computer simulation of high length chain polymers, we show that simple excluded volume interactions between polymer and nanoparticles lead to a wealth of changes in the diffusion coefficient and entanglement density of the chains. This opens up the possibility of using nanoparticles for tuning polymer properties, such as toughness, melt viscosity, and transient rubberlike behavior. © 2010 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 48: 687–692, 2010     

Grafting of maleic anhydride onto linear polyethylene: A Monte Carlo study

Monte Carlo simulation was used to study the graft of maleic anhydride (MAH) onto linear polyethylene (PE‐g‐MAH) initiated by dicumyl peroxide (DCP). Simulation results revealed that major MAH monomers attached onto PE chains as branched graft at higher MAH content. However, at extremely low MAH content, the fraction of bridged graft was very close to that of branched graft. This conclusion was somewhat different from the conventional viewpoint, namely, the fraction of bridged graft was always much lower than that of branched graft under any condition. Moreover, the results indicated that the grafting degree increased almost linearly to MAH and DCP concentrations. On the other hand, it was found that the amount of grafted MAH dropped sharply with increasing the length of grafted MAH, indicating that MAH monomers were mainly attached onto the PE chain as single MAH groups or very short oligomers. With respect to the crosslink of PE, the results showed that the fraction of PE‐(MAH)_n‐PE crosslink structure increased continuously, and hence the fraction of PE‐PE crosslink decreased with increasing MAH concentration. Finally, quantitative relationship among number average molecular weight of the PE, MAH, and DCP contents was given. © 2004 Wiley Periodicals, Inc. J Polym Sci Part A: Polym Chem 42: 5714–5724, 2004     

Rheological properties of HyperMacs—long‐chain branched analogues of hyperbranched polymers

HyperMacs are long chain branched analogues of hyperbranched polymers, differing only in the sense that they have polymer chains, rather than monomers between branch points. Although the building blocks for HyperMacs and AB₂ macromonomers can be well defined in terms of molecular weight and polydispersity, the nature of the coupling strategy adopted for the synthesis of the HyperMacs results in branched polymers with a distribution of molecular weights and architectures. Melt rheology showed polystyrene HyperMacs to be thermorheologically simple, obeying William–Landel–Ferry behavior. Zero shear viscosities of the polymers were shown to increase with average molecular weight and the melts display shear‐thinning behavior. HyperMacs showed little evidence for relaxation by reptation and the rheological behavior agreed well with the Cayley tree model for hierarchical relaxation in tube models of branched polymers. © 2007 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 45: 2762–2769, 2007     

Monte Carlo simulation of size exclusion chromatography for randomly branched and crosslinked polymers

The elution curves of size exclusion chromatography for nonlinear polymers formed through random branching and crosslinking of long polymer chains were simulated with a Monte Carlo method. We considered two types of measured molecular weight distributions (MWDs): (1) the MWD calibrated relative to standard linear polymers and (2) the MWD obtained with a light scattering (LS) photometer in which the weight‐average molecular weight of polymers within the elution volume is determined directly. The calibrated MWDs clearly underestimate the molecular weights for both randomly branched and crosslinked polymers, and this technique can be used to assess the degree of deviation from the true MWD. When the primary chains conform to the most probable distribution, the calibrated MWD can be estimated reasonably well with the Zimm–Stockmayer equation for the g factor with the help of the relationship between the average number of branch points per molecule and the degree of polymerization. However, the LS method gives good estimates of the true MWD for both randomly branched and crosslinked polymers, although the agreement is better for the branched ones. © 2000 John Wiley & Sons, Inc. J Polym Sci B: Polym Phys 38: 2009–2018, 2000     

Monte Carlo simulation of miscibility of polymer blends with repulsive interactions: effect of chain structure

The effects of the chain structure and the intramolecular interaction energy of an A/B copolymer on the miscibility of the binary blends of the copolymer and homopolymer C have been studied by means of a Monte Carlo simulation. In the system, the interactions between segments A, B and C are more repulsive than those between themselves. In order to study the effect of the chain structure of the A/B copolymer on the miscibility, the alternating, random and block copolymers were introduced in the simulations, respectively. The simulation results show that the miscibility of the binary blends strongly depends on the intramolecular interaction energy () between segments A and B within the A/B copolymers. The higher the repulsive interaction energy, the more miscible the A/B copolymer and homopolymer C are. For the diblock copolymer/homopolymer blends, they tend to form micro phase domains. However, the phase domains become so small that the blend can be considered as a homogeneous phase for the alternating copolymer/homopolymer blends. Furthermore, the investigation of the average end-to-end distance () in different systems indicates that the copolymer chains tend to coil with the decrease of whereas the of the homopolymer chains depends on the chain structure of the copolymers. As for the system containing the alternating or the random copolymers, the homopolymer chains also tend to coil with the decrease of . However, for the systems including the block copolymers, there is a slight difference in the of the homopolymer chains with the variation of .     

Revisiting Chain Transfer to Polymer and Branching in Controlled Radical Polymerization of Butyl Acrylate

Acrylic monomers undergo chain transfer to polymer during polymerization leading to branched and even gelled polymers. It has been experimentally demonstrated that the extent of branching is higher for conventional free radical polymerization than for controlled radical polymerization (ATRP, RAFT, NMP) and this has been qualitatively explained in terms of the differences in the concentrations of highly reactive short‐chain radicals between controlled and conventional radical polymerizations. Contrary to this explanation, in this work, it is quantitatively demonstrated that the short transient lifetime of the radicals, i.e., the time between activation and deactivation of the radical in controlled radical polymerization, is the cause for the low level of branching in these polymerizations.

     

Elastic behavior of uniform star polymer chains

Elastic behaviors of uniform star polymer chains with two to seven branches (namely, functionality f = 2-7) are investigated using Monte Carlo simulation and the bond fluctuation model. Here chain dimensions and thermodynamic properties of uniform star polymer chains during the process of tensile elongation are studied, and comparisons with linear chain are also made. Static properties of chains such as chain sizes and asphericities of chains are calculated, and g-value of g = 〈S²〉_star/〈S²〉_linear decreases with elongation ratio increasing for different functionality f. Thermodynamic properties such as average energy 〈U〉, free energy per bond 〈A′〉 and elastic force F are also investigated during the process of tensile elongation. In the meantime, scatting functions P(q) are calculated for star polymer chains with different functionality f. Additionally, we also discuss the influence of elongation ratio on scattering form factor. The impenetrability of the star cores is known to cause a discontinuity in the osmotic pressure showed through a peak in the scattering functions, and some different behaviors in the tensile process for uniform star chain are obtained.     

Wavelet-accelerated Monte Carlo sampling of polymer chains

We introduce a new method for coarse-graining polymer chains, based on the wavelet transform, a multiresolution data analysis technique. This method, which assigns a cluster of particles to a coarse-grained bead located at the center of mass of the cluster, reduces the complexity of the problem significantly by dividing the simulation into several stages, each with a small fraction of the number of beads in the overall chain. At each stage, we compute the distributions of coarse-grained internal coordinates as well as potential functions required for subsequent simulation stages. We show that, with this wavelet-accelerated Monte Carlo method, coarse-grained Gaussian and self-avoiding random walks can reproduce results obtained from atomistic simulations to a high degree of accuracy in orders of magnitude less time. © 2005 Wiley Periodicals, Inc. J Polym Sci Part B: Polym Phys 43: 897–910, 2005     

Rheological Characterization of Computationally Synthesized Reactor Populations of Hyperbranched Macromolecules: Bivariate Seniority‐Priority Distribution of ldPE

Summary: Hyperbranched molecules like low‐density polyethylene (ldPE) adopt a huge variety of molecular architectures. Previous work has shown that it is possible to computationally synthesize these architectures and to characterize them according to radius of gyration. Here, a method is presented and applied on ldPE to characterize populations using rheological quantities in terms of comb‐shaped and Cayley tree structures. Interbranch segments are assigned seniority and priority values that quantify their behavior in relaxation and elastic deformation processes. New general‐purpose algorithms have been developed to derive the full bivariate seniority/priority distribution using a representation from the graph theory of branched architectures. This paper describes the computation of bivariate chain length/degree of branching distributions (CLD/DBD) using a Galerkin finite element method for two scission mechanisms: linear and topological scission. The DBD is calculated using pseudo‐distributions. Random scission is treated with fragment length and branch point redistribution functions as obtained from scission statistics of branched molecules, preferentially yielding short and long fragments. Reactor populations of ldPE architectures are then obtained using computational synthesis. The seniority and priority distributions calculated indeed prove to be an adequate characterization method. They show good comparison, although not a complete overlap, with size characterization using a variant of the radius of gyration. It was possible to calculate a full bivariate seniority/priority fraction distribution, but due to the limited sample size its surface was not smooth. Subsequent work has shown the consequences for the prediction of rheological properties.

Seniority/priority values for segments of molecules for one chain length/number of branch points combination.