Carrier Transport and Gas‑Sensing Performance of Asymmetric MoS₂ Schottky Diodes via Charge‑Transfer‑Induced Barrier Modulation

Background

Since the discovery of graphene, a plethora of 2D materials have emerged, each offering distinct electronic and mechanical characteristics. Transition metal dichalcogenides (TMDs), such as MoS₂ and WSe₂, are especially attractive because they provide a finite bandgap that graphene lacks, enabling transistor operation with high on/off ratios and significant carrier mobilities (≈200 cm² V⁻¹ s⁻¹ for monolayer MoS₂) [1–5]. Nonetheless, realizing MoS₂ transistors that rival silicon devices requires addressing challenges related to lattice integrity, fabrication techniques, and, critically, contact resistance between the metal electrodes and the MoS₂ channel.

Most prior work has focused on symmetric junctions and largely relied on band‑structure analysis alone. However, such approaches do not capture the dynamic carrier behavior under gas exposure, which is governed by changes in the SBH at the metal/MoS₂ interface. By incorporating both experimental measurements and first‑principles calculations, this study bridges that gap and provides a quantitative understanding of how different metal contacts influence gas‑induced barrier modulation in MoS₂ Schottky diodes.

Method

Fabrication of MoS₂ Devices

Few‑layer MoS₂ flakes were mechanically exfoliated from bulk crystals (SPI supplies) and transferred onto heavily doped Si/SiO₂ substrates using a PDMS stamp (Sylgard 184). Platinum and aluminum electrodes, each 100 nm thick, were defined by electron‑beam lithography (JEOL JSM‑7001F) and subsequently deposited by electron‑beam evaporation. Device performance was evaluated at room temperature by measuring source–drain and source–gate characteristics with a Keithley 2400 source meter.

Surface Potential Measurement

The surface potential was mapped using Kelvin probe force microscopy (KPFM) in interleave mode on a Veeco Nanoscope IV. A PtIr‑coated silicon probe (SCM‑PIT) was first used to capture topography, then to record the electrostatic force signal, enabling extraction of the work‑function difference between the metal contacts and the MoS₂ channel.

DFT Calculations

DFT simulations were performed with VASP using the PBE functional and van der Waals corrections (D2). A 3×3 supercell of monolayer MoS₂ (three Mo atoms, six S atoms) was constructed on six‑layer Al(111) and Pt(111) slabs. After full geometry optimisation, the band structures and work functions were extracted. Gas‑molecule adsorption studies included NO₂ and NH₃; the resulting charge transfer and SBH shifts were quantified for each metal contact.

Result and Discussion

Atomic force microscopy confirmed a MoS₂ thickness of 4 nm. KPFM revealed a markedly lower work‑function difference between Al and MoS₂ (≈22 %) compared to Pt and MoS₂ (≈100 %), indicating that Al forms an almost Ohmic contact while Pt forms a pronounced Schottky junction.

Current–voltage (I‑V) measurements showed that the Al‑contacted device produced a linear, high‑current drain response, exceeding that of the Pt device by more than three orders of magnitude, confirming the lower SBH of Al. Transfer characteristics under forward bias (0.1 V, 5 V, 10 V) exhibited typical n‑type behavior; the Al device achieved on/off ratios of ≈10⁶ at 5 V, whereas the Pt device remained around 10³. Threshold voltage analysis via √I_DS–V_G plots placed the Al device near –70 V and the Pt device near –30 V, again underscoring the lower barrier of Al.

DFT results corroborated these findings: the calculated SBH for Al/MoS₂ was 0.16 eV, while for Pt/MoS₂ it was 0.44 eV—a 72 % reduction attributable to the lower Al work function. Band‑structure plots (Figure 3) displayed the Fermi level lying closer to the conduction band in the Al case, explaining the enhanced carrier injection.

When asymmetric Al/Pt electrodes were introduced, rectifying behavior emerged, with current flowing preferentially from Al to Pt. The asymmetric configuration further amplified the influence of the Al contact on carrier transport, as evidenced by a pronounced shift in threshold voltage when the source‑drain bias was varied.

Real‑time gas‑sensing experiments measured the relative resistance change (ΔR/R_air) under 3 V bias for NOₓ and NH₃ at 10, 20, and 30 ppm. NOₓ, acting as an electron acceptor, increased the device resistance due to p‑doping and SBH elevation. Conversely, NH₃, an electron donor, reduced resistance by lowering the SBH. Sensitivity rose with concentration for both gases, but NOₓ consistently produced a stronger response, consistent with the DFT‑predicted SBH shifts of 0.16 eV (NO₂) versus 0.13 eV (NH₃).

These experimental observations align with theoretical predictions, reinforcing the concept that gas‑induced charge transfer at the metal/MoS₂ interface is the primary driver of sensor response. The combination of asymmetric contacts and a tunable SBH offers a pathway to high‑performance, selective gas detectors.

Conclusion

The study demonstrates that metal contact engineering—specifically using Al for its low work function—significantly reduces the Schottky barrier in MoS₂ devices, enhancing current flow and enabling robust gas sensing. KPFM and DFT analyses jointly confirm that higher‑work‑function metals, such as Pt, increase SBH, while lower‑work‑function metals lower it. NOₓ exposure markedly raises device resistance through charge withdrawal, whereas NH₃ lowers resistance via electron donation, mirroring the calculated SBH variations. These findings highlight the critical role of asymmetric contacts in designing MoS₂‑based sensors with tailored sensitivity and selectivity.

Abbreviations

AFM

Atomic force microscopy

DFT

Density functional theory

FET

Field‑effect transistor

KPFM

Kelvin probe force microscopy

SBH

Schottky barrier height

TMDs

Transition metal dichalcogenides

V_ds

Source‑drain voltage

vdW

Van der Waals