The application of reverse Monte Carlo modelling to a polymeric melt

The reverse Monte Carlo (RMC) modelling technique has been applied to the polymeric melt poly(propylene oxide) (PPO). The modelling was based on neutron diffraction experiments on both hydrogenous and deuterated samples. Inter-atomic distances and calculated bond angle distributions for the RMC produced model were found to be in good agreement with reported experimental results, showing the ability of the RMC method to produce realistic structural models of amorphous polymers on the molecular length scale. The strengths and limitations of the RMC models in studies of the intermediate and long range order in polymers are investigated with emphasis on the information content and information quality. Particularly, we find by studying the RMC model of PPO that the strong first diffraction peak at about 1.45 Angstrom(-1) is almost entirely due to weak inter-chain correlations. The local chain conformation was investigated by calculating the partial atomic pair correlation functions for atoms belonging to monomers close in sequence. The results show that the most probable conformation is a "stretched" trans conformation, where two consecutive methyl groups are pointing in almost opposite directions. Although the dihedral and bond angle distributions found in the model are probably somewhat broader than in reality, they show unambiguously interpretable features which support the findings above. (C) 2000 Elsevier Science Ltd. All rights reserved.