Hydrophobic Polymer Encapsulation Effects on Subgap Density of States in Multilayered Molybdenum Disulfide Field‐Effect Transistors

Abstract

Herein, exfoliated, multilayered molybdenum disulfide (MoS2) (m‐MoS2) field‐effect transistors (FETs) are implemented with bilayered SiNx/SiOx gate dielectrics on indium tin oxide (ITO) substrates. For a quantitative understanding on gas adsorption effects on the electrical performance of m‐MoS2 FETs, subgap density of states (DOSs) in m‐MoS2 layers without (or with) hydrophobic polymer encapsulation are extracted using optical charge‐pumping capacitance–voltage spectroscopy. Based on extracted subgap DOSs and their deconvolution with analytical model of acceptor (or donor) like states, all electrical parameters are systematically analyzed. More importantly, two times increase in field‐effect mobility (μFE) is strongly related with decrease in shallow donor states (NSD) from 2 × 1018 to 2 × 1017 eV−1 cm−3. In addition, significant improvement of subthreshold swing (SS), hysteresis gap (VHYS) are attributed to the reduction of tail states (NTA) from 4 × 1019 to 2 × 1019 eV−1 cm−3, along with decrease in midgap states (NMid) from 3 × 1016 to 1.3 × 1016 eV−1 cm−3. For a final validation, technology computer‐aided design (TCAD) simulation with extracted DOS information nicely replicate measured I–V characteristics for m‐MoS2 FETs without (or with) encapsulation, indicating that extracted DOS information is quite accurate, compared with implemented m‐MoS2 FETs.

Unscrambling for Subgap Density-of-States in Multilayered MoS2Field Effect Transistors under DC Bias Stress via Optical Charge-Pumping Capacitance-Voltage Spectroscopy

“…In the present work, a wavelength of 1065 nm was intentionally selected in consideration of the energy bandgap of m-MoS2 layers because lasers with wavelengths shorter than 1065 nm have a higher chance of exciting electrons located in the valence band toward conduction via band-to-band generation, leading to overestimation of the density of trapped charges. Moreover, optical power was selected in this work at a condition of 5 mW after confirming the saturation behaviors on photoresponse, which is similar to the method reported in the literature [28].…”

“…(2)), respectively, yielding information for a quantitative understanding. In particular, it is generally reported that the μFE in n-channel FETs is significantly affected by defects near EC [28]. Several key parameters, such as gTA(E), gSD(E), and gDA(E), below Ec are energetically distributed throughout the energy bandgap with energy location dependency, yielding capture and thermal release of free carriers in certain energy states, such as multiple trapping and thermal release events [28].…”

“…In this sense, identification of the subgap density of states (subgap-DOSs) in TMDCs and their correlation between operating conditions and device reliability are indispensable to the improvement of electrical parameters as well as device reliability in MoS2 field effect transistors (FETs), which is a prerequisite for reliable operation in real applications. Thus, there have been significant activities to investigate the initial state of material properties via ab initio calculations and density of state characterization via optical charge pumping, Terman methods, and multifrequency C-V characterization [28][29][30]. However, there have been limited reports on detailed subgap-DOSs and their evolution for the corresponding bias stress for MoS2 FETs, and moreover, the origin of subgap-DOSs and their variation associated with both external gas ambients (e.g., O2, H) and internal atomic structure deformation remain elusive even after tremendous research activity in this field [28][29][30][31][32][33][34][35][36][37].…”

“…Thus, there have been significant activities to investigate the initial state of material properties via ab initio calculations and density of state characterization via optical charge pumping, Terman methods, and multifrequency C-V characterization [28][29][30]. However, there have been limited reports on detailed subgap-DOSs and their evolution for the corresponding bias stress for MoS2 FETs, and moreover, the origin of subgap-DOSs and their variation associated with both external gas ambients (e.g., O2, H) and internal atomic structure deformation remain elusive even after tremendous research activity in this field [28][29][30][31][32][33][34][35][36][37]. This is possibly due to several hurdles: (i) the inability to prepare almost identical device ensembles associated with the absence of a mature, scalable process scheme for the preparation of active layers, (e.g., MoS2 layers) with both high crystallinity and wafer-scaled substrates, (ii) limited scanning [29], multifrequency method for near conduction bands [30], and hysteretic gate transfer characteristics for deep level traps [31].…”

“…However, there was a systematic report to characterize the full range of subgap-DOSs for m-MoS2 FETs by using the optical charge pumping method, which has been popularly utilized and studied for the investigation of the reliability of a-IGZO TFTs with full maturity up to the commercialization level in AMOLED panels [38][39][40][41][42][43]. Interestingly, entire investigations on device reliability and its evolution on subgap-DOSs, corresponding to positive (or negative) bias stress (PBS or NBS), are completely missing, albeit specifically reported for initial subgap-DOSs for m-MoS2 FETs with (or without) hydrophobic polymer (e.g., CYTOP) encapsulation [28].…”