A Computational Analysis of Crystallite Shape under Quiescent and Stretch-Induced Polyethylene Crystallization

One of the most remarkable properties of crystallizable polymers is their ability to form complex semicrystalline morphologies, which are responsible for their excellent barrier and very good mechanical properties. These morphologies are sensitive to various processing and material-dependent factors such as the temperature, the flow type and strain rate, the molecular architecture and the size distribution of polymer chains; their prediction is challenging. Atomistic simulations can offer unique insight into the mechanisms that govern polymer crystallization, elucidating aspects of nucleation and growth under different processing conditions. In this work, starting from Monte Carlo equilibrated linear polyethylene melts of a uniform molecular weight distribution, we perform isothermal molecular dynamics simulations of crystallization under two different protocols, i.e., under (a) quiescent and (b) stretching conditions. Comparing between the two protocols, we quantify how the presence of a flow field affects the emerging semicrystalline morphology. Using home-built algorithms, we calculate the evolution of the degree of crystallinity over time, analyze the mass and the radius of gyration tensor of the largest ordered cluster present, and determine the stochastic distribution of induction times from the aforementioned geometric characteristics and from mean first passage time analysis. We show that the presence of a flow field has strong impact on nucleation and growth, accelerating the emergence of the crystalline phase. We establish a methodology for quantifying the shape and orientation of the emerging crystallites and of their constituent atoms through a radius of gyration tensor and Q-tensor analysis and highlight differences in their evolution between the two different crystallization protocols. Crystallites created under stretching are strongly oriented along the drawing direction and ultimately adopt a more cylindrical symmetry, as opposed to the quasi-spherical clusters generated under quiescent conditions, whose orientation is random. We quantifyvia the eigenvalues of the diagonalized radius of gyration tensorthe growth rate along the principal axes of the crystallites, as well as their volumetric growth rate. Finally, we investigate our quiescent specimens at long times, presenting evidence of semicrystalline morphologies such as bridging between different crystallites and curving of the lamellar planes. We showcase, throughout our study, links between crystallite morphology and the stages of nucleation and growth, shedding additional light on the mechanisms which govern polymer crystallization. All of our findings are accompanied by three-dimensional visualizations of the systems under study, illuminating the parameters and mechanisms analyzed, as well as the influence of stretching on the orientation of the chains participating in the crystalline phase. We validate the various parameters and measures obtained through our different methods of analysis against each other.