ALD/ALE 2022 Session NS-WeM2: 2D Materials I

Layered chalcogenide materials with van der Waals bonding between layers are fascinating materials. Some are semiconductors with bandgaps suitable for transistors or have been used in atomically thin memory devices, while others are thermoelectrics or topological insulators that are interesting for quantum devices. Antimony(III) telluride has a layered structure with van der Waals bonds between Te atoms in adjacent quintuple layers. We have performed ALD of Sb₂Te₃ using bis(trimethylsilyl)tellurium and antimony(III) ethoxide, which is a pair of previously reported precursors,¹ expanding the set of studied substrates and employing various surface treatments prior to deposition. Growth of a continuous layer is highly dependent on the starting surface through at least the first 500 cycles. Growth on the van der Waals surface of epitaxial monolayers of MoS₂ was difficult, as has been the case for other thermal ALD processes on transition metal dichalcogenides.² On SiO₂, we observed incompletely coalesced islands by SEM. Conversely, on amorphous GeTe, single-crystal Bi₂Te₃, or polycrystalline TiN substrates subjected in situ toa remote Ar plasma, we grew continuous Sb₂Te₃. On amorphous GeTe and TiN, we measured a pronounced drop in sheet resistance after 500 cycles. X-ray photoelectron spectroscopy confirms the presence of Sb and Te with 3d_5/2 core levels near 527 eV and 572 eV, respectively. Compared to reference spectra, the ratio of intensities matches the expected stoichiometry. Bi₂Te₃ crystals are interesting because they have the same structure as Sb₂Te₃. In our prior study of plasma enhanced ALD on MoS₂, we observed that growth on single-crystal GaN, which has nearly the same lattice parameter in the basal plane, led to much less out-of-plane growth than on other substrates.³In reports on ALD of Sb₂Te₃ with the same Te precursor but antimony trichloride instead of antimony(III) ethoxide,⁴ or growth of Sb with analogous chloride and alkylsilyl precursors,⁵ engineering the hydroxyl groups on the substrate seems to play an important role in promoting in-plane nucleation and rapid coalescence. However, our experiments with bis(triethylsilyl)tellurium and antimony(III) ethoxide did not show an analogous advantage. We will further consider atomic layer epitaxy of van der Waals solids in this presentation. We thank ONR N0014-18-12511 for support and the NSF 2DCC DMR-1539916 for epitaxial substrates.

1. Eom et al., Chem. Mater. 24, 2099, 2012

2. Walter et al., Appl. Surf. Sci. 480, 43, 2019

3. Mughal et al., J. Vac. Sci. Technol. A 37, 010907, 2019

4. Nminibapiel et al., ECS J Solid State Sci Technol 3P95, 2014

5. Mohney et al., ALD2016

2D transition metal dichalcogenides (TMDs) have many unique properties such as high electronic mobilities and a large surface-to-volume ratio, making them promising materials for various applications such as nanoelectronics and electrocatalysis. In particular, alloys of 2D TMDs hold great potential due to their composition-controlled properties, making them more versatile than pure TMDs. Atomic layer deposition can be used to synthesize thin films of such TMD alloys with sub-monolayer growth control, making this method particularly relevant for the aforementioned applications. In this work we employ plasma-enhanced atomic layer deposition (PEALD) to synthesize alloys of the 2D TMDs NbS₂ and WS₂, i.e. Nb_xW_1-xS₂ using W(NtBu)₂(NMe₂)₂ and Nb(NtBu)(NEt₂)₃ as precursors and H₂S plasma as coreactant. We present a thorough characterization of the alloy growth, showing excellent composition control ranging from nanocrystalline WS₂ to amorphous NbS₂ based on ellipsometry, XPS, SEM, Raman spectroscopy and TEM. The morphology of the films depends strongly on the composition, as the alloy films have a significantly enhanced occurrence of out-of-plane oriented crystallites.

The suitability of Nb_xW_1-xS₂ as a channel material for transistors is evaluated by Hall measurements and FET device characterization. We find that the incorporation of only 10% Nb reduces the resistivity of the WS₂ film by 4 orders of magnitude at a mobility of 0.3 cm²/Vs. Furthermore, improving the Nb distribution in the film through consecutive precursor dosing (ABC-type cycles) further increases the mobility by a factor of 2.5. At the same time we observe that the formation of out-of-plane oriented crystallites is strongly suppressed when using this process type, which is critical for applications in nano-electronics. Fabrication and characterization of FETs based on the alloy films are currently ongoing. Secondly, we assess the efficacy of the Nb_xW_1-xS₂ films as electrocatalysts for the hydrogen evolution reaction (HER). The alloys were found to outperform the pure NbS₂ and WS₂ films deposited by PEALD, having lower Tafel slopes and overpotentials. The alloy with a composition x = 0.84 performed best with a Tafel slope of 131 mV/dec and an overpotential of 470 mV.

Our results show that the Nb_xW_1-xS₂ alloys improve on the electrical and electrocatalytic performance of pure WS₂ and NbS₂. Furthermore, the various supercycle schemes provide a valuable platform for fundamental insight into the surface reactions and growth mechanics during the PEALD process and for further fine-tuning of the alloy functionalities.

2D semiconductor transition metal dichalcogenides have attracted considerable attention due to their layered structure, suitable band gap, electrochemically active unsaturated edges and relatively good stability against photocorrosion. These properties result promising for different applications including, Li-ion batteries, photocatalysis and hydrogen evolution reaction (HER). Apart from the widely studied 2D MoS₂, 2D selenide and telluride equivalents, MoSe₂ and MoTe₂,have recently gained considerable interest due to their higher electrical conductivity, wider inter-layer distance and narrower bandgap as compared to MoS₂, high surface area and close to zero Gibbs free energy edges for hydrogen adsorption.

Unlike sulfide dichalcogenides, the lack of Se and Te precursors have prevented the synthesis of selenide and telluride dichalcogenides by ALD. In order to surpass such impediment, we present a set of novel in-house synthesized Se and Te compounds, which were successfully combined with commercial Mo precursor to synthesize MoSe₂ and MoTe₂ by ALD [1-5]. The as-deposited ALD MoSe₂ and MoTe₂ on substrates of different nature were extensively characterized by different techniques, which confirmed the chemical composition and revealed the growth of 2D flaky nano-crystalline MoSe₂ and MoTe₂. In parallel, MoSe₂ and MoTe₂@TiO₂ nanotube layers (TNTs) heterostructures were fabricated in a simple and fast fashion to explore and exploit the MoSe₂ and MoTe₂ photo- and electrocatalytic properties. TNTs act as excellent photoactive supporting material providing a high surface area, unique directionality for charge separation, and highly effective charge collection.

The presentation will introduce and describe the synthesis of the 2D Mo dichalcogenides, the corresponding physical and electrochemical characterization and encouraging results obtained in HER [4,5], photocatalysis [4-6] and Li-ion batteries [7].

[1] R. Zazpe et al, FlatChem (2020) 21 100166

[2] J. Charvot et al, Chempluschem (2020) 85 576

[3] J. Charvot et al, RSC Adv. (2021) 11 22140

[4] R. Zazpe et al, ACS Appl. Nano Mater. (2021) 3 12 12034

[5] R. Zazpe et al, Appl. Mater. Today (2021) 23 101017

[6] M. Motola et al, Nanoscale (2019) 11 23126

[7] H. Sopha et al FlatChem (2019) 17 100130

Layered two-dimensional molybdenum sulfide (MoS₂) has attracted great interest for a promising candidate material for opto-electronics and photo sensors applications due to its unique characteristics such as tunable bandgap, high electron mobility and high current on/off ratio. In order to apply MoS₂ to the industrial field, significant efforts have been placed in obtaining a wafer-scale uniform MoS₂. Plasma-enhanced atomic layer deposition (PEALD) is a promising approach for depositing a 2D MoS₂ because of its excellent thickness control. However, the low growth temperature of PEALD makes it difficult to guarantee the quality of the MoS₂ thin films.