Analysis of the conductive behavior of a simplified sediment system and its computational simulation

Abstract

Because of the important contributions of electrochemical redox reactions to biochemical cycles and their potential application for the in-situ remediation of contaminated sediment, the mechanisms of long-distance electron transport coupling spatially separated redox half reactions in sediment have drawn much attention. To explore a preliminary mechanism of long-distance electron transport in sediment, in the current study, two simplified composite systems are constructed consisting of spherical ferroferric oxide (Fe3O4) nanoparticles and rod-like carbon nanotubes (CNTs) as conductive fillers and silica (SiO2) particles as the matrix. Two different constructed composite systems (e.g., SiO2/Fe3O4 and SiO2/Fe3O4/CNTs) were used to model a three-dimensional sediment framework instead of sediment with quite complex components. The effects of the loading of conductive fillers (e.g., Fe3O4, CNTs) and the particle size of SiO2 matrix on the conductive behavior of the composite system were investigated. The results showed that both of the electrical properties of SiO2/Fe3O4 and SiO2/Fe3O4/CNTs composite systems typically exhibited a non-linear conductive behavior that the electrical conductivity increased with the increasing of filler loading and showed an abrupt increase at critical filler loading. The conductivity of the SiO2/Fe3O4 and SiO2/Fe3O4/CNTs composite systems with micro-sized SiO2 as the matrix was higher than that of the composite systems with nano-sized SiO2 as the matrix. Compared with the SiO2/Fe3O4 composite system, the electrical conductivity of the SiO2/Fe3O4/CNTs composite system was enhanced by several orders of magnitude and only a small loading of CNTs could make the conductivity of the SiO2/Fe3O4/CNTs composite system reach a higher level. The electrical conductivity predicted by the electrical conductivity model of a two-phase composite system showed a similar trend as the experimental results and the two-dimensional (2D) percolation-based model filled with rods gave a good estimation of percolation probability.