Spectroscopic Characterization of Process Dependent 2D Material Devices

<p dir="ltr">As Moore's law encourages the reduction in size of transistors and computing chips, a shift in the architecture of logic devices is necessitated. This shift suggests the eventual limit of silicon, even as it was once an ideal semiconducting material. To solve this problem, two-dimensional transition metal dichalcogenides provide a potential alternative for the future of ultrascaled devices. \ce{MoS_2}, in particular, is a promising possible alternative as an n-type transistor material to silicon. Realizing larger application of these new two-dimensional devices, however, requires optimizing their performance, efficiency, and reliability. Optimization, in turn, requires knowledge of the physical mechanisms that drive electrical response. Identifying these mechanisms necessitates characterizing them in different material stacks and with differing fabrication techniques. Spectroscopic techniques including Raman and photoluminescence spectroscopy probe such changes. They can thus be used to show the variation and induced changes between material stacks, providing insight on the physics of device-to-device variation and failure. </p><p><br></p><p dir="ltr">In this work, spectroscopy is utilized to quantify material damage, changes in carrier concentration and strain, and variations in doping between MoS₂ transistors fabricated in differing architectures and process steps. The differences quantified by spectroscopy, in turn, are correlated with variations in device characteristics, including threshold voltage, subthreshold slope, mobility, and current. It is shown that device improvement and degradation can be explained using a variety of material characteristics identified through spectroscopic techniques. First, it is shown that seed layers play a critical role in the charge transfer doping of 2D MoS₂ devices. Second, we show that HIM FIB processing induces measurable damage to MoS₂ nanoribbons, and it is possible to use spectroscopy to quantify improvements via capping. Together, this work highlights how optical spectroscopy is a necessary tool to continue the development and improvement of 2D logic devices. </p>