Design of electrically and thermally conductive polymer composites for electronic packaging

In designing electrically and thermally conductive polymer composites, one must know the fundamental knowledge of their macroproperty-microstructure relations. The macroproperties of interest in this study are the effective electrical and thermal conductivities. The key microstructural factors include filler shape, size, size distribution and geometric arrangement of fillers.; The main tasks consist of processing, characterization and analytical modeling of silver flake/polymer matrix composites. Polymer composites are processed in various silver flake concentrations. Characterization includes the cure analysis, the microstucture examination, and the measurement of electrical and thermal conductivities. Four analytical models are constructed for predicting the effective electrical and thermal conductivities; (1) three-dimensional percolation model, (2) three-dimensional electrical resistor network model, (3) two-dimensional bond percolation model, and (4) three-dimensional thermal resistor network model. A three-dimensional percolation model predicts the threshold volume fraction of flake in terms of flake aspect ratio. A three-dimensional resistor network model is developed for the effective electrical resistivity, based on percolation and microscopic conduction mechanisms. It is found that flakes with larger aspect ratio, smaller mean value in size and broader size distribution yield smaller threshold volume fraction. The analytical predictions reasonably agree with the experimental results of silver flake/polymer matrix composites. A two-dimensional bond percolation model predicts the effect of filler arrangement on the composite resistivity. A composite with segregated distribution of conducting particles yields much smaller threshold volume fraction, comparing with random distribution. The analytical predictions are well matched with measured results of SiC particle/Si{dollar}sb3{dollar}N{dollar}sb4{dollar} composites. A three-dimensional thermal resistor network model is developed for the effective thermal conductivity, where high thermal conductivity ratio of filler to matrix and direct contacts among flakes are expected. The analytical predictions show a good agreement with measured data of silver flake/polymer matrix composites.; Finally a three-dimensional percolation model is developed for studying the piezoresistive behavior of Ni-coated graphite short fiber/elastomer matrix composites, realizing that the microstructure of conducting short fibers can be changed under finite deformation. It is found that an initially conductive composite becomes non-conductive around a critical strain, exhibiting a switching behavior. The analytical predictions agree well with the experimental data.