Semiconductor/fluoride structures on Si for 2D electronics

Nowadays, silicon technology is everywhere, starting with phones and other gadgets for daily use up to extremely powerful supercomputers and server farms. The impressive success of silicon technology over the past 50 years has relied on scaling, that is, by continuously shrinking the devices from generation to generation to make them more affordable and faster. Semiconductor devices typically consist of stacked insulating and semiconducting layers and in these scaled devices some layers have become only a few atoms thin and difficult to scale any further. In addition, these layers have different lattice constants, which makes it difficult to stack them without introducing problems in the transition region. One promising solution is the use of 2D materials, which are atomically thin materials that form chemical bonds only in a two-dimensional plane, and are held together by van der Waals forces in the third dimension. Theoretically, these materials can be arbitrarily stacked without having to worry about lattice mismatch, resulting in clean transition layers and crystalline material stacks. In practice, however, there are many fabrication challenges. Also, when considering the targeted application range of 2D devices, namely ultrascaled nanoelectronics, we foresee problems with the currently used insulators, as none of those appear to be suitable when scaled down to an required effective thickness below one nanometer. Various alternative options are currently considered, but none of them have as of yet resulted in devices competitive with silicon technology. We have recently demonstrated that the use of fluorides as insulators results in very stable and scalable 2D devices. The use of fluorides would have many advantages because they are ionic crystals with low defect densities, which are excellent insulators, their growth processes are reasonably well understood compared to any other crystalline insulator for 2D technologies, and, most importantly, their inert F- terminated surface forms van der Waals interfaces with 2D materials. In this project we will theoretically investigate the properties of fluorides when used as ultrascaled insulators for 2D devices. We will calculate their properties and electronic structure using ab initio calculations and then determine to what thickness they can be scaled without resulting in unacceptable leakage currents. In addition, we will determine the most likely to occur point defects and how they impact the device behavior when fluorides are combined with typically used 2D semiconductors like MoS2 . Our calculations will be benchmarked against back-gated prototype devices with CaF2 grown on Si substrates on which exfoliated 2D semiconductors will be transferred. These calculations will provide estimates on which material combinations result in the best device performance.