| Since graphene was found in 2004, the applications of two-dimensional (2D) materials have attracted wide attention to academia, due to unique electronic and optoelectronic properties. Transition metal dichaldogenide (TMDC), which has strong planar covalent bond and weak van der Waals interactions between adjacent sheets, is the MoS2 popular type of 2D materials beyond graphene. Mechanical exfoliation is common method used to separate individual sheets from bulk 2D layered crystals by breaking the weak van der Waals bonds between the layers. Although research of TMDC has decades of history, due to the characterization and fabrication process of 2D materials developed in recent years, applications on the devices arise a lot of new opportunities.Molybdenum disulfide (MoS2) and tungsten disulfide (WS2) are typical wide bandgap (with bandgap of 1.8 eV and 2.0 eV in monolayer samples) 2D semiconductor, which is a great precondition for low static power and high on/off ratio. The 2D structure can effectively restrain short channel effect, induced by devices scaling down. Also, the process of TMDC devices is compatible with tranditional silicon-based devices. Mobility has been regarded as one of the important parameters for device performance. Althouth phonon-limited mobility of monolayer MoS2 (WS2) at room temperature is 410 cm2/Vs (over 1000cm2/Vs), according to theoretical calculation, the experimental value before was only 40～50 cm2/Vs. The prior work in our group has found that there are a large number of atomic defects in natural MoS2, which restricted the improvement of device performance.Also, the transport mechanism for MoS2 is not very clear, and we cannot quantificationally explain observations in the experiment, such as metal to insulating transition. In view of current condition, we have developed a series of study in MoS2 and WS2 transport. The main contents are as followed:(1) Through statistical analysises of high resolution transmission electron microscopy (HR-TEM) images before and after repairment, we show that sulfur vacancies (SV), which is the main type of intrinsic defects in MoS2, can be effectively repaired by (3-mercaptopropyl)trimethoxysilane (MPS) under mild annealing. We investigated three kinds of samples, untreated-Top side treated (TS-Treated)-Double sides treated (DS-Treated), to study the effect of defect repairment to charge transport. Surprisingly, record-high mobility greater than 80cm2/Vs is achieved in DS-treated MoS2 field-effect transistors at room temperature. By fitting experiment data, we find that the reason for MIT in monolayer MoS2 is localized states in the bandgap induced by defects. Furthermore, we develop a theoretical model to systematically understand the mobility, conductivity, and metal-insulation transition in monolayer MoS2.(2) We deposited ～10nm Al2O3, with dielectric constant of 10, on the 300nm SiO2/Si substrates by atomic layer deposition (ALD). Then we fabricated the monolayer WS2 devices on the high-K substrates. We found that mobility at room temperature was improved 100% comparing with bare SiO2 and metallic transport was observed at the higher carrier density. We have fitted our experimental results through the similar model we have used in MoS2, and found that the trap density reduced significantly on high-κ dielectrics, resulting in a great improvement in device performance. Moreover, we found that thiol chemistry passivation can effectively reduce the density of interface traps and Coulomb impurities, leading to a significant improvement of carrier mobility and the transition of charge transport from the insulating to the metallic regime. Record-high mobility of 83cm2/Vs (337cm2/Vs) is reached at room temperature (low temperature). Furthermore, we develop a theoretical model incorporating Coulomb impurities and traps to explain the observed mobility improvement. |
PDF Full Text Request