Answers are grounded in this paper and cite their sources.
Tungsten diselenide (WSe2) is the leading p-type channel candidate for two-dimensional complementary field effect transistors (CFETs). However, depositing high-k gate dielectric on its inert basal plane has still been a challenge. Modified atomic layer deposition (ALD) chemistries to avoid unintentional doping, plasma surface treatments, and interfacial seed layers are among a few promising methods for high-k integration. Extensive studies have been done on molybdenum disulfide (MoS2) to identify possible routes to high-performance NFET. This study presents a possible route to integrating high-k HfO2 on salt-assisted atmospheric pressure chemical vapor deposition (SA-CVD)-grown WSe2 on Si/SiO2 wafers. We compare evaporated Hf, Ti, and Ge seed layers, oxidized in ambient, prior to high-k ALD on SA-CVD WSe2. Because salt-assisted growths bind defect density to local precursor flux, plasma-based surface treatments can damage or modify the underlying channel material, making seed layers a favorable option over plasma processes. ALD without any seed layer suppresses the hole branch of both exfoliated and CVD WSe2 transistors, where peak drive currents were lowered by roughly six orders of magnitude. Seedless HfO2 ALD, Hf and Ti seeds eliminate hole conduction below 0.5 nA/μm. Germanium is the only one of the three seeds that we study that preserves the hole branch, at least at the level of an unseeded reference device at equal gate overdrive, and shows a +40 V threshold voltage shift, making the device hole dominant. Therefore, Ge is a promising seed for retaining hole conduction in WSe2 PFETs, and the seed-ALD high-k HfO2 is a possible dielectric stack for top-gated PFETs.
No takes yet. Share an insight, caveat, or question.
Singh et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: