The effect of water absorption on the dielectric properties of epoxy nanocomposites

Abstract

In this research, the influence of water absorption on the dielectric properties of epoxy resin and epoxy micro-composites and nano-composites filled with silica has been studied. Nanocomposites were found to absorb significantly more water than unfilled epoxy. However, the microcomposite absorbed less water than unfilled epoxy: corresponding to the reduced proportion of the epoxy in this composite. The glass transition temperatures (T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">g</sub> ) of all the samples were measured by both differential scanning calorimetry and dielectric spectroscopy. The T <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">g</sub> decreased as the water absorption increased and, in all cases, corresponded to a drop of approximately 20 K as the humidity was increased from 0% to 100%. This implied that for all the samples, the amount of water in the resin component of the composites was almost identical. It was concluded that the extra water found in the nanocomposites was located around the surface of the nanoparticles. This was confirmed by measuring the water uptake, and the swelling and density change, as a function of humidity as water was absorbed. The water shell model, originally proposed by Lewis and developed by Tanaka, has been further developed to explain low frequency dielectric spectroscopy results in which percolation of charge carriers through overlapping water shells was shown to occur. This has been discussed in terms of a percolation model. At 100% relative humidity, water is believed to surround the nanoparticles with a thickness of approximately 5 monolayers. A second layer of water is proposed that is dispersed but sufficiently concentrated to be conductive; this may extend for approximately 25 nm. If all the water had existed in a single layer surrounding a nanoparticle, this layer would have been approximately 3 to 4 nm thick at 100%. This "characteristic thickness" of water surrounding a given size of nanoparticle appeared to be independent of the concentration of nanoparticles but approximately proportional to water uptake. Filler particles that have surfaces that are functionalized to be hydrophobic considerably reduce the amount of water absorbed in nanocomposites under the same conditions of humidity. Comments are made on the possible effect on electrical aging.