Low-frequency noise in multilayer MoS₂field-effect transistors: the effect of high-k passivation

Abstract

Diagnosing of the interface quality and the interactions between insulators and semiconductors is significant to achieve the high performance of nanodevices. Herein, low-frequency noise (LFN) in mechanically exfoliated multilayer molybdenum disulfide (MoS2) (~11.3 nm-thick) field-effect transistors with back-gate control was characterized with and without an Al2O3 high-k passivation layer. The carrier number fluctuation (CNF) model associated with trapping/detrapping the charge carriers at the interface nicely described the noise behavior in the strong accumulation regime both with and without the Al2O3 passivation layer. The interface trap density at the MoS2-SiO2 interface was extracted from the LFN analysis, and estimated to be Nit ~ 10(10) eV(-1) cm(-2) without and with the passivation layer. This suggested that the accumulation channel induced by the back-gate was not significantly influenced by the passivation layer. The Hooge mobility fluctuation (HMF) model implying the bulk conduction was found to describe the drain current fluctuations in the subthreshold regime, which is rarely observed in other nanodevices, attributed to those extremely thin channel sizes. In the case of the thick-MoS2 (~40 nm-thick) without the passivation, the HMF model was clearly observed all over the operation regime, ensuring the existence of the bulk conduction in multilayer MoS2. With the Al2O3 passivation layer, the change in the noise behavior was explained from the point of formation of the additional top channel in the MoS2 because of the fixed charges in the Al2O3. The interface trap density from the additional CNF model was Nit = 1.8 × 10(12) eV(-1) cm(-2) at the MoS2-Al2O3 interface.

Phonon scattering limited performance of monolayer MoS2 and WSe2 n-MOSFET

“…[1][2][3] Apart from monolayer flakes of such MX 2 materials, few layers thick MX 2 MOSFET have also been the focus of various device fabrication studies. [7][8][9] The role of electron-phonon scattering in such MX 2 layers do have a significant impact on the performance of such 2-D channel MOSFETs. [10][11][12] Recently Guo et al have shown the effect of phonon scattering in monolayer MoS 2 MOSFETs.…”

Controlled Low‐Frequency Electrical Noise of Monolayer MoS2 with Ohmic Contact and Tunable Carrier Concentration

“…FETs are shown by the light brown asterisk (PE-free devices) and the red asterisk (PE-encapsulated devices). Exfoliated monolayer [15] and multilayer [36] MoS 2 from the literature are the light cyan and blue shadows, respectively. Representative values of α H in exfoliated [39] and CVD [32] graphene (denoted by the wine shadow and red shadow, respectively) are also shown for comparison.…”

“…The McWhorter model suggests that the trapping-detrapping dynamics of the interfacial traps result in the 1/f noise, [13,28] and gives S g e k TD C WLf where D it is the interfacial trap density, k B is the Boltzmann constant, and T is the temperature. [17,36] The transition of the S I amplitude from the carrier density dependent (i.e., S I ∝ N) to the carrier density independent (i.e., S I = constant) may be well explained by the variation of the interfacial trap density in the PE: D it increases with V eg at V eg < −0.4 V and gets to 1.6 × 10 13 cm −2 eV −1 in the linear I D -V eg region, which is about www.advelectronicmat.de two orders of magnitude greater than the trap density in the SiO 2 interface. [17,36] An alternative method to determine D it from the subthreshold swing (see Section S.4, Supporting Information) yields D it = 2.2 × 10 13 cm −2 eV −1 , in good agreement with the value extracted from the McWhorter model.…”

“…[17,36] The transition of the S I amplitude from the carrier density dependent (i.e., S I ∝ N) to the carrier density independent (i.e., S I = constant) may be well explained by the variation of the interfacial trap density in the PE: D it increases with V eg at V eg < −0.4 V and gets to 1.6 × 10 13 cm −2 eV −1 in the linear I D -V eg region, which is about www.advelectronicmat.de two orders of magnitude greater than the trap density in the SiO 2 interface. [17,36] An alternative method to determine D it from the subthreshold swing (see Section S.4, Supporting Information) yields D it = 2.2 × 10 13 cm −2 eV −1 , in good agreement with the value extracted from the McWhorter model. Figure 3b displays 1/A and I D 2 against V eg (the data in Figures 2e and 1c) for comparison.…”

Percolative switching in transition metal dichalcogenide field-effect transistors at room temperature

“…Similar behavior for both WSe 2 and MoS 2 FETs imply We chose WSe 2 FETs as the primary experimental platform due to the following reasons: First, WSe 2 is an emerging TMDC FET material with several desirable properties ranging from high carrier mobility due to low effective mass of the carriers, ambipolar conduction and superior chemical stability compared to sulphides [12,13,28,32,33,35,37]. Second, in spite of the progress in standard electrical transport properties [19,[28][29][30][31][32][33][34][35], the origin and magnitude of intrinsic 1 f -noise, a crucial performance limiting factor in electronic device applications, in WSe 2 FETs is not known, and third, given the recent studies of noise in MoS 2 FETs [46][47][48][49][50], similar studies in WSe 2 allows identification of the generic aspects of noise processes in TMDC FETs, which in turn, provides crucial insight to microscopic details of charge distribution and disorder.…”