Conductivity relaxation in the interfacial phase of iron core-iron oxide shell nanocomposites

Abstract

The alternating current electrical conductivity of a gel-derived glass of composition 55Fe₂O₃·45SiO₂ (mol%) was measured over a frequency range of 100 Hz to 6 MHz. The gel was subjected to a reduction treatment at 923 K for ½ h and subsequently heated in ordinary atmosphere at a temperature 773 K for ½ h to grow a Fe-core Fe₃O₄ shell nanostructure with a median diameter of 6.2 nm. This formed a percolative network within the silica gel. Mossbauer spectra confirmed the presence of Fe₃O₄ in the nanoshell. Electrical measurements were also carried out on these nanocomposites at different frequencies and temperatures. Isothermal electrical modulus data for both reference and treated glass systems were analyzed using both the CK0 and CK1 Kohlrausch-related frequency response models. Reference-glass shape parameter values, estimated by fitting the experimental data to the K0 model at several temperatures, were found to be ~0.32. Here, the K0 model led to much better fits than the K1 did. However, for the treated core-shell-structured nanocomposite material, both models yielded good fits with consistent but different shape parameter estimates: very close to ½ for the K0 model and ⅓ for the K1 model. In accordance with the structural measurements and with axiomatic topological considerations that predict a shape-parameter value of ⅓ for one-dimensional motion and ½ for two-dimensional motion, it appears that the ~0.32 value is consistent with one-dimensional motion of charge carriers along the narrow channels of the interconnected iron-rich three-dimensional phase of the reference glass. Further, although the K1-model ⅓ estimates for the treated material also indicate the presence of one-dimensional charge motion at the two-dimensional interface between the two interconnected phases of the reference glass, the ½ K0 estimates for the same material suggest an effective charge-motion dimension of 2. Importantly, comparison of the high-frequency dielectric constant estimates for the K0 reference glass and the K1 treated one clearly leads to the new but physically plausible conclusion that the bulk frequency-independent dielectric constant of about 30 is independent of the treatment.