Epoxy-graphite Resistive Composites

In this thesis, we proposed to study polymer-based thick-film resistors (TFRs) made of graphite particles dispersed in an epoxy matrix. Three main axes were investigated: formulation of the composites, characterization with the determination of electrical, mechanical and thermo-mechanical properties of the material, potential applications. Special care was first given to the formulation of the composite formed by an epoxy matrix in which graphite particles are dispersed, providing conductivity to the material. As application temperatures of a polymer is limited by its glass transition Tg we used two epoxies: a standard Tg one (ca. 90 °C) and a high Tg one (ca. 200 °C). Particle size and shape effects were also investigated: synthetic graphite (ellipsoidal shape) and expanded graphite (worm-like shape) were taken as fillers. For comparative purposes, carbon black was also tested. For TFRs to be deposited by the standard process of screen-printing, viscosity of the paste must be controlled. Therefore, we performed a study on solvents and we proposed an appropriate set adapted to epoxies. Functionalization of the particle surface through an oxidation process was studied. The aim was to create reactive groups, which could further react with the polar group of the epoxy and provide a better adhesion. The main part concerned the characterization of these materials, which helps to understand their behavior. Electrical properties were first determined for our composites. We showed that particle size has little effect on the resistivity, whereas a specific surface area and a high structure of the filler give significantly lower resistivity values at low volume fractions. As our composites should form percolative systems, we fit the results using the theoretical conductivity formula. However, the values determined were incoherent, thus, we think that the conductivity in our composites is driven by the so-called "debris model", i.e. the overall conductivity depends on the small debris dispersed in the matrix. This was confirmed by the microscopic analysis of our composites. The piezoresistivity study showed that larger particles tend to give smaller gauge factors, which is linked to the aspect ratio of the particles. In these experiments, temperature and substrate effects highlighted the importance of the Tg regarding stability of the properties. Dynamic Mechanical Analysis (DMA) was then performed on our materials. As expected, the higher the volume fraction, the higher the storage modulus. Larger particles and expanded graphite gave the same tendency, due to the fact that, overall, "less" resin was present in the composite. A better adhesion between epoxy matrix and functionalized particles was also demonstrated: oxidation led to a higher storage modulus. In parallel to DMA, thermal expansion of the composites was determined with an optical dilatometer. A one-time shrinkage was observed in the standard epoxy, which highlighted a "cycling effect" of the epoxy. The shrinkage was found to decrease with the filler volume fraction, and was not observed with expanded graphite, these two parameters diminishing the amount of resin in the composite. On the contrary, the high Tg epoxy did not show this effect. We proposed that this is linked to its two-step polymerization process. The values of Tg were determined from DMA and dilatometry and were additionally compared to the values obtained with Differential Scanning Calorimetry (DSC). Same tendencies were found for all methods, proving the consistency between our results. Finally, these composites were used in different applications. Firstly, combined with sacrificial layers, they allowed the fabrication of microstructures such as microchannels, or cantilevers. Appropriate formulations based on sublimable polyols were developed, the high advantage being the fabrication of fine structures, by avoiding capillary forces linked with evaporation of the liquid phase. However, strong interactions between the sacrificial layers and the epoxy occurred. Over-layers based on ethylcellulose and silicone loaded with synthetic graphite were therefore formulated. The high potential of the process was demonstrated. Secondly, composite material formulations for the development of a testing set-up for the oil industry were investigated. The aim is to recreate a dummy and yet representative rock to help in the quality control of tools that measure the resistivity. The use of composites based on polymer and carbon-based filler presents several advantages: large range of available resistivities, ease of the process and transportable set-up. Comparative tests between TFRs and bulk objects were made. Differences in resistivity appeared and were found to come from the affinity of the epoxy with water, the amount of water trapped inside a bulk object being higher and more difficult to evaporate because of the thickness. Therefore, in order to manufacture more massive objects than thick-films, a longer polymerization time should be envisaged to allow the diffusion of water molecules from the material.