Search

Thermal interface material (TIM) is pivotal for the heat dissipation between layers of high-density electronic packaging. The most widely used TIMs are particle-filled composite materials, in which highly conductive particulate fillers are added into the polymer matrix to promote heat conduction. The numerical simulation of heat transfer in the composites is essential for the design of TIMs; however, the widely used finite element method (FEM) requires large memory and presents limited computational time for the composites with dense particles. In this work, a numerical homogenization algorithm based on fast Fourier transform was adopted to estimate the thermal conductivity of composites with randomly dispersed particles in 3D space. The unit cell problem is solved by means of a polarization-based iterative scheme, which can accelerate the convergence procedure regardless of the contrast between various components. The algorithm shows good precision and requires dramatically reduced computation time and cost compared with FEM. Moreover, the effect of the particle volume fraction, interface thermal resistance between particles (R-PP), interface thermal resistance between particle and matrix (R-PM), and particle size have been estimated. It turns out that the effective conductivity of the particulate composites increases sharply at a critical filler volume fraction, after which it is sensitive to the variation of filler loading. We can observe that the effective thermal conductivity of the composites with low filler volume fraction is sensitive to R-PM, whereas the it is governed by R-PP for the composites with high filler content. The algorithm presents excellent efficiency and accuracy, showing potential for the future design of highly thermally conductive TIMs.

We show here successful large-area growth of triangular-shaped ZnO nanoplate arrays on silicon substrates and the investigation of their electron field emission properties. The special geometric-shaped nanoplate emitters have a very large surface area for heat dissipation and large field enhancement factors. These lead to very stable and low-threshold field emission. The fluctuations of the macroscopic current density are less than 1.5%. The threshold macroscopic field is about 4.3 V mu m(-1) at an emitter-anode gap of 525 mu m. The stable electron field emission is ascribed to low local heating effects, originating from the superior heat dissipation ability and the low Joule heating of the sheet-like emitters. Plotting Fowler-Nordheim curves yields straight lines with small deviations at low and high macroscopic fields. The origins of these deviations are discussed. Our experiments demonstrate a promising design for stable and efficient emitters.

P-type CuO nanorods were synthesized by a hydrothermal method and the ethanol-sensing properties of sensors based on CuO were investigated. The sensor resistance increased when it was exposed to ethanol and decreased in the air, which is contrary to the case for sensors realized from n-type semiconductor. The resistance of the CuO-based sensor was about 2 kΩ in air and 6 kΩ in ethanol vapour with concentration of 2000 ppm. Such a sensing property is attributed to surface accumulation conduction. Sensors based on CuO nanorods have potential applications in detecting ethanol in low concentration.

Fast humidity sensors are reported that are based on CeO2 nanowires synthesized by a hydrothermal method. Both the response and recovery time are about 3 s, and are independent of the humidity. The sensitivity increases gradually as the humidity increases, and is up to 85 at 97% RH. The resistance decreases exponentially with increasing humidity, implying ion-type conductivity as the humidity sensing mechanism. A model based on the morphology and surface energy of the nanowires is given to explain these results further. Our experimental results indicate a pathway to improving the performance of humidity sensors.