Dominic Waldhör Dipl.-Ing. BSc Publications

Biography Dominic Waldhör was born in Linz, Austria in 1992. He studied at the TU Wien, Vienna, Austria where he received the BSc degree in Electrical Engineering and the Diplomingenieur degree in Microelectronics and Photonics in 2018. He joined the Institute for Microelectronics in October 2018 where he works as PhD student. His current research focuses on the ab initio simulation of oxide defects in semiconductor devices.

CaF₂, or calcium fluoride, has demonstrated significant potential as a dielectric material in emerging two-dimensional (2D) electronic devices. Its wide bandgap of E_g=12.1 eV substantially reduces gate-leakage currents, when compared to other insulators commonly used for 2D devices, such as hexagonal boron nitride. Moreover, its chemically inert surface, resulting from the strong ionic bonding, allows for the formation of quasi-van der Waals interfaces with various 2D materials, including graphene and transition metal dichalcogenides (TMDs). A possible gate stack featuring CaF₂ grown on an Si(111) substrate is illustrated in Fig. 1.

Despite its advantages as a gate dielectric, CaF₂-based devices still exhibit greater hysteresis during gate voltage sweeps and a larger bias temperature instability (BTI) compared to established silicon technologies, currently employed in the semiconductor industry. Both of these phenomena are generally caused by charge trapping at defect sites in the insulator. In this work, we conduct a theoretical study based on density functional theory (DFT) to identify the possible defects which are responsible for BTI and hysteresis in CaF₂-based devices. Specifically, we focus on silicon-impurity defects, as CaF₂ is typically grown on a Si(111) substrate by molecular beam epitaxy (MBE) for device applications. During this process, individual Si atoms from the initial Si surface can be incorporated into the CaF₂ film at interstitial sites (Si_i) or replace a calcium atom (Si_Ca), as depicted in Fig. 2.

Utilizing DFT, we calculated the relaxation energy and the thermodynamic trap levels – the most crucial parameters governing charge trapping dynamics – for the Si_i and Si_Ca defects. We discovered that both defects can exist in multiple charge states, ranging from -2 to +4, with trap levels across the CaF₂ bandgap. As illustrated in Fig. 3, depending on the 2D semiconductor used, these trap levels can be energetically close to either the valence or conduction band, and thus can potentially act as electrically active charge trapping centers. Since the investigated defects are intrinsic to CaF₂ grown on Si(111), they may be the limiting factor for device reliability in this material system.

Fig. 1: Si(111)/CaF₂/MoS₂ gate stack. Due to its good lattice match, CaF₂ can be grown on Si(111) in thin layers. It provides a virtually defect-free quasi van der Waals interface to the 2D material and hence greatly improves the reliability of 2D devices.

Fig. 2: Projected density of states (PDOS) and localized highest occupied molecular orbitals of the Si_i (top) and Si_Ca (bottom) defect. Both defects introduce localized states in the bandgap, as can be seen from the DOS projected onto the Si impurity. Note that the Si PDOS is scaled by a factor of 50 for illustration purposes.

Fig. 3: Trap levels of the Si_i and Si_Ca defects within the bandgap of CaF₂, together with the band edges of silicon and several commonly used 2D semiconductors. Alignments between band edges and active defect states are indicated with a dashed line.