In materials that are subject to environmental reactions, one way to increase durability is to decrease the chemical interdiffusion rate. In metal systems, mass transport by diffusion is enhanced – if any change – with finer grained microstructures, because of the greater number of high diffusion interface and boundary paths. In non-metals, increased transport is not necessarily the case, as other mechanisms may control the overall diffusivity. C. Wagner [1] showed the theoretical basis of transport in certain nanocomposite cases based on space charge layers at interfaces.

In an oxide-in-oxide nanocomposite, one can use the tools of bulk analysis to study the near-interface effects on transport. The immiscible pair, TiO2-SiO2, was used as a model system. Nanophase powders (50 and 200 m2/g, respectively) from flame hydrolysis were dispersed with high shear, then freeze-dried and hot-pressed to near theoretical density. Transmission electron microscopy was used to show < 10 nm particles in the composites produced at the time of diffusivity measurement. The rate of change of four-point electrical conductivity in response to a change in oxygen partial pressure was used to estimate chemical diffusivity. The re-equilibration differences between TiO2 diffusivity for bars of about 10 times different square cross-section areas confirmed the interpretation of the conductance rate changes to be the result of bulk interdiffusion.

The interdiffusion rate in the nanoscale dispersoid composites was more than 5 times lower than for same SiO2-fraction composites after coarsening which were like TiO2 with no second-phase. Results are in qualitative agreement with models for impurity segregation to interfaces and for space charge layers near the dispersed nanoparticles.