Our era is defined by its technology, and our future is dependent on its continued evolution. Over the past few decades, we have witnessed the expansion of advanced technology into all walks of life and all industries, driven by the exponential increase in the speed and power of semiconductor-based devices.

However, as the length scale of devices reaches the atomic scale, a deep understanding of atomistic theory and its application is increasingly crucial. In order to illustrate the power of an atomistic approach to understanding devices, we will present results and conclusions from three interlinked projects: n-type doping of III-nitride semiconductors, defects for quantum computing, and macroscopic simulations of devices.

First, we will study effective n-type doping of III-nitride semiconductors and their alloys, and analyze the barriers to effective n-type doping of III-nitrides and their alloys. In particular, we will study the formation of DX centers, and predict alloy composition onsets for various III-nitride alloys. In addition, we will perform a comprehensive study of alternative dopants, and provide potential alternative dopants to improve n-type conductivity in AlN and wide-band-gap nitride alloys.

Next, we will discuss how atomic-scale defects can act as a curse for the development of quantum computers by contributing to decoherence at an atomic scale, specifically investigating the effect of two-level state defects (TLS) systems in alumina as a source of decoherence in superconducting qubits based on Josephson junctions; and also as a blessing, by allowing the identification of wholly new qubits in different materials, specifically showing calculations on defects in SiC for quantum computing applications.

Finally, we will provide examples of recent calculations we have performed for devices using macrosopic device simulations, largely in conjunction with first-principles calculations. Specifically, we will discuss the power of using a multi-scale approach to accurately model oxide and nitride-based heterostructures, and thereby illustrate our ability to predict device performance on scales unreachable using a purely first-principles approach.