Development of a ceramic-based composite for direct bonded copper substrate

• Published in: Ceramics international 2017-04, Vol.43 (6), p.5236-5246
• Main Authors: Akhtar, S.S., Kareem, L.T., Arif, A.F.M., Siddiqui, M.U., Hakeem, A.S.
• Format: Article
• Language: English
• Subjects: Alumina; Ceramics; Composites; Computational; Design; Direct bonded copper substrate
• ISSN: 0272-8842; 1873-3956
• DOI: 10.1016/j.ceramint.2017.01.049

In the present paper, a computational approach is presented to design alumina-based composite with tailored properties that could replace commercial alumina used in Direct Bonded Copper (DBC) substrates for applications in power electronic modules. A mean-field homogenization and effective medium approximation (EMA) using an in-house code is used for predicting potential optimum thermal and structural properties for DBC substrates by considering the effect of filler type, volume, and size in the alumina matrix. The primary goal for designing such alumina-based composites is to have enhanced thermal conductivity for effective heat dissipation and spreading capabilities together with a coefficient of thermal expansion (CTE) value that is close to the silicon chips in electronic circuits in order to avoid interface layers. At the same time, other functional properties like elastic modulus and electrical conductivity have to be maintained. Our strategy incorporates thermal and structural properties of composites as a constraint on the design process. Among various metallic and carbon-based fillers, chromium, silicon carbide and diamond fillers were found suitable candidates that could enhance the thermal and structural performance of the alumina-based substrates. As a validation, we developed alumina-silicon carbide (Al2O3-SiC) composites in line with the designed range of filler size and volume fraction using Spark Plasma Sintering (SPS) process. Thermal and structural properties including thermal conductivity, CTE, and elastic modulus are measured to complement the computational design. It is found that the developed computational design tool is accurate enough in predicting the desired properties of composite materials for DBC substrate applications.