The structure and the percolation behavior of a mixture of carbon nanotubes and molecular junctions: a Monte Carlo simulation study.

The structure and the percolation behavior of the composite of carbon nanotubes (CNTs), CNT molecular junctions and polymers are studied using Monte Carlo (MC) simulations. We model a CNT as a rigid rod composed of hard spheres. "X" and "Y" molecular junctions of CNTs are constructed by joining four and three segments of CNTs, respectively. The model system consists of CNTs mixed with either "X" or "Y" molecular junctions. The system is equilibrated using Monte Carlo simulations and the equilibrated configurations are used to locate the clusters of connected molecules via a recursive algorithm. The fraction (P(perc)) of configurations with a percolating cluster is then estimated for a given total volume fraction (phi(t)) of molecules. When P(perc) reaches 0.5, phi(t) of the system is considered a percolation threshold concentration (phi(c)). The percolation behavior is found to be sensitive to the aspect ratio of CNTs and the concentration and the shape of molecular junctions. phi(c) is decreased with an increase in the aspect ratio of CNTs. As the mole fraction of molecular junctions is increased, phi(c) is decreased significantly, which suggests that molecular junctions could enhance the electric conductivity of CNT-polymer composites. X junctions are found to construct a percolating network more effectively than Y junctions. More interestingly, even though molecular junctions change the percolation behavior significantly, the site-site pair correlation functions of CNTs hardly show any difference as the mole fraction of molecular junctions is increased. This implies that the percolation of CNTs is determined by the subtle many-body correlation of CNTs that is not captured by the site-site pair correlation functions.