Computational materials engineering: Capabilities of atomic-scale prediction of mechanical, thermal, and electrical properties of microelectronic materials

Atomic-scale computational materials engineering offers an exciting complement to experimental observations, revealing critical materials property data, and providing understanding which can form the basis for innovation. This contribution reviews the current state of atomic-scale simulations and their capabilities to predict mechanical, thermal, and electric properties of microelectronics materials. Specific examples are the elastic moduli of compounds such as aluminum oxide, the strength of an aluminum/silicon nitride interface, the first-principles prediction of coefficients of thermal expansion of bulk aluminum and silicon nitride, thermal conductivity of polyethylene, the prediction of the diffusion coefficient of hydrogen in metallic nickel, the calculation of dielectric properties of zinc oxide and optical properties of silicon carbide. The final example illustrates the control of the work function in the HfO
₂
/TiN interface of a CMOS gate stack. For an increasing number of materials properties, computed values possess accuracies similar to measured data. Such accuracy has become possible due to advances in theoretical approaches and numerical algorithms combined with the astounding increase in compute power.