Patterned, low-temperature growth of transition metal dichalcogenides for low resistance raised contacts
Abstract
Transition metal dichalcogenide (TMD) monolayers are promising channel materials for next-generation electronic devices. A challenge is the high contact resistance between monolayer TMDs and metal contacts, especially for holes. In this regard, raised source/drain contacts are promising. However, the direct, patterned growth of raised contacts at CMOS-compatible temperatures remains largely unresolved. We present plasma-free selenization and sulfurization of metal-oxides at substrate temperatures down to 400 °C, compatible with back-end-of-line (BEOL) thermal budgets. To achieve growth at such temperatures, gas-phase chalcogen precursors are first thermally activated at 950 °C. Films grown on single-crystal monolayer TMDs exhibit high crystal quality as confirmed by transmission electron microscopy. Raised contacts on WSe2 monolayers fabricated using this approach yield a low hole contact resistance of 0.3 kΩ·μm after chemical doping. This process is shown to be applicable to growing various TMDs, including WS2, MoSe2, MoS2, PtSe2, and NbS2.
Dataset DOI: 10.5061/dryad.b5mkkwht8
Description of the data and file structure
Data for the main figures of the paper entitled "Patterned, Low-Temperature Growth of Transition Metal Dichalcogenides for Low Resistance Raised Contacts". Simulated mole fractions of selenium allotropes (Fig. 1b), Raman signal from grown WSe2 films based on different activation temperatures (Fig. 1e). Raman signal of WSe2 grown from various substrate temperatures (Fig. 2a). X-Ray Diffraction pattern of selenized WOx films (Fig. 2b). Raman signal from various transition metal dichalcogenide films grown with our low-temperature growth method (Fig. 4b, Fig. 4c, and Fig. 4d). Transfer length method data for various lengths of short channel device (Fig. 5b and Fig. 5c). Top-gate device characteristics (Fig. 6b and 6c).
Files and variables
File: Files.zip
Description: Each folder corresponds to the figure number in the main text.
Folder 1:
Se_concentrations.xlsx has the simulated mole fraction of each selenium allotrope (Se, Se2, ..., Se8) using CANTERA, shown in Fig. 1b. Each sheet has the name of the allotrope (ex, Se2 for Se2). Unit of the x-axis is in °C (Temperature of the simulation). Unit of the y-axis is the mole fraction of the allotrope existing (dimensionless).
Raman.xlsx has the Raman signal of WSe2 grown at 400 °C activation temperature (Fig. 1e in grey) in the sheet named '400 (Grey).' Raman signal of WSe2 grown at 950 °C activation temperature (Fig. 1e in red) in the sheet named '950 (Red).' Unit of the x-axis is in 1/cm (Wavenumber). Unit of the y-axis is the count of photons during the Raman measurement.
Folder 2:
Raman_Temperatures.xlsx has all the Raman signals of WSe2 grown at different substrate temperatures (140 °C, 200 °C, 340 °C, and 410 °C) shown in Fig. 2a. Each sheet inside the file corresponds to the substrate temperature. Unit of the x-axis is in 1/cm (Wavenumber). Unit of the y-axis is the normalized count of photons during the Raman measurement.
XRD.xlsx has x-ray diffraction data for the selenized tungsten oxide film shown in Fig. 2b. The unit of the x-axis is in degrees (2θ). The unit of the y-axis is the count of photons during the X-ray diffraction measurement.
Folder 4:
Raman_TMDs.xlsx has all the Raman data from various transition metal dichalcogenides other than WSe2 grown for the paper (WTe2, WS2, MoSe2, MoS2, PtSe2, and NbS2) in Fig. 4. Each sheet name corresponds to the transition metal dichalcogenide. Unit of the x-axis is in 1/cm (Wavenumber). Unit of the y-axis is the normalized count of photons during the Raman measurement.
Folder 5:
TLM.xlsx has the transfer length measurement data from various channel length devices (0.21 μm, 0.36 μm, 0.62 μm, 0.82 μm, and 1.1 μm) of exfoliated monolayer WSe2. Sheets named with numbers (0.21, 0.36, 0.62, 0.82, and 1.1) have the current value (in μA/μm) at each biasing drain bias (in V). The sheet named 'TLM' has the values for Figure 5c, where column A has each length of the devices (0.21 μm, 0.36 μm, 0.62 μm, 0.82 μm, and 1.1 μm), with column B having the corresponding total resistance of the device in kΩ∙μm. Column D has the x-axis value (channel length in μm) and the corresponding fitted total resistance in kΩ∙μm in Column E.
Folder 6:
Fig6b.xlsx has the top-gate device characteristics based on the top-gate bias of the device. Sheets named as 'V_DS=-4 V' and 'V_DS=-3 V' have the current values (in μA/μm) according to different top-gate biases (in V) when the drain bias is -4 V and -3 V, respectively. Sheet named I_TG has the top-gate leakage current values (in μA/μm) according to different top-gate biases (in V).
Fig6c.xlsx has the top-gate device characteristics based on the drain bias of the device. Sheets named as 'V_TG=-4 V,' 'V_TG=-2 V,' and V_TG=-1 V have the current values (in μA/μm) according to different drain biases (in V) when the top-gate is biased at -4 V, -2 V, and -1 V, respectively.