The semiconductor industry’s ever-growing demand for smaller device dimensions, better power efficiency and lower prices has pushed state of the art silicon-based computing technology to its physical limits. Ferroelectric non-volatile technology could be the answer to this challenge as it enables memory and logic tasks to be performed in the same physical units. This could lead to in-memory computing applications that are vastly energy efficient over present-day von Neumann architectures. My thesis focuses on the opportunities and challenges posed by 2D/3D heterojunctions in logic and memory devices. In this defense, I demonstrate sub-10 nm thick ferroelectric Al_1-xScₓN films on Al and Sc into capacitors, diodes and transistors with 2D TMDC channels that reach sub-2 V operation. I demonstrate that ultrathin Al_1-xScₓN can maintain robust ferroelectricity even when grown on industry-standard substrates that are compatible with present-day back-end-of-line processes. I also highlight the potential of Al_1-xScₓN beyond low-power operation, by demonstrating the resilience of Al_1-xScₓN devices to gamma radiation—making this material also applicable to space and nuclear industries. Finally, I present the anisotropic effects of non-ferroelectric wide-bandgap semiconductor Ga2O3 on 2D/3D diode performance, and the implications for prospective industrial uses of Ga₂O₃. Overall, my work aims to explore scaling and reduced power consumption in next-generation electronics, and to encourage further research and innovation in this rapidly evolving field.