Why Three Layers of MoS2 Make the Perfect Electrical Contact

For more than a decade, engineers chasing the promise of atomically thin electronics have run into the same stubborn obstacle: the moment a metal electrode touches a two-dimensional semiconductor, resistance spikes at the junction and bleeds away much of the performance that made ultrathin materials so attractive in the first place. A new study published in PLOS One by Luhao Liu, Zhengyi He, Yixiang Li, Shengsheng Lin and Songlin Li now adds a surprising twist to that familiar problem. Measuring how electrical contact resistance changes as molybdenum disulfide, one of the most studied two-dimensional semiconductors, is stacked layer by layer, the team found that the relationship is anything but straightforward. Resistance does not simply fall or climb with thickness. Instead it dips to a pronounced minimum at exactly three atomic layers and then rises again, a non-monotonic pattern the authors trace to a tug-of-war between two competing physical mechanisms deep inside the crystal.

The material at the center of the work, MoS2, belongs to a family of compounds that can be peeled into sheets just three atoms thick while retaining useful semiconducting behavior. Unlike graphene, which conducts electricity superbly but lacks an intrinsic bandgap, MoS2 offers a genuine gap between its valence and conduction bands, making it a serious candidate for transistors that could push past the scaling limits of silicon. But a transistor is only as good as its contacts, and in two-dimensional devices the metal–semiconductor junction often dominates the total resistance. When a metal such as antimony is deposited on MoS2, charge carriers must cross an energy barrier known as the Schottky barrier, whose height depends delicately on the band structure of the semiconductor at the interface. Any factor that reshapes that band structure, including how many layers of the material are stacked together, can therefore move the performance needle dramatically.

To quantify this layer dependence systematically, the researchers fabricated field-effect transistors built on MoS2 channels of varying thickness, from a single atomic layer up to many layers, each fitted with antimony contacts. Rather than relying on a single device measurement, which can be skewed by fabrication quirks, they extracted contact resistance using the transmission line method, a well-established technique in which identical devices are patterned with channels of several different lengths. By measuring the total resistance of each device and plotting it against channel length, the extrinsic contribution from the contacts separates cleanly from the intrinsic resistance of the MoS2 channel itself, allowing the junction resistance to be isolated with confidence. The team repeated this analysis for devices built on two different gate dielectrics, silicon dioxide and hafnium dioxide, to check that the observed trends were robust properties of the MoS2–antimony interface rather than artifacts of one particular substrate chemistry.

The resulting dataset revealed the striking non-monotonic behavior. Moving from a monolayer to a trilayer, contact resistance steadily decreased, bottoming out at 308 ohm-micrometers for the three-layer devices, a figure that represents a substantial improvement over the monolayer case. But the improvement did not continue. As the researchers stacked on additional layers beyond three, the contact resistance reversed course and began climbing again, erasing the gains made in the few-layer regime. The fact that the same pattern appeared on both silicon dioxide and hafnium dioxide dielectrics strengthens the conclusion that the minimum at three layers reflects intrinsic contact physics rather than some accident of device processing. For device engineers, the message is unusually concrete: when optimizing a MoS2 transistor with antimony contacts, the thickness of the channel is not a free parameter to be tuned casually, because the sweet spot sits at a specific, countable number of atomic layers.

Explaining why three layers occupy the optimum required the authors to dig into the electronic band structure of few-layer MoS2 and the way charge carriers actually move through a stacked crystal. In a monolayer, the bandgap is at its largest, roughly in the range of 1.8 to 1.9 electronvolts in direct-gap form, and as layers are added the gap narrows and the band character shifts. That narrowing matters enormously at a metal contact, because the height of the Schottky barrier that electrons must surmount is tied directly to the position of the conduction band edge relative to the metal’s Fermi level. A smaller bandgap pulls the band edge downward, lowering the barrier and letting carriers flow more easily across the junction. In the few-layer regime, this bandgap reduction effect works in favor of the contact, and it explains the steady drop in resistance observed between one and three layers.

But the same stacking that shrinks the bandgap also introduces a penalty. In a multilayer crystal, carriers injected at the top surface, where the metal contact sits, must hop or tunnel between adjacent layers to traverse the material and reach the conduction pathway. Each interlayer hop adds its own resistance, and the more layers there are, the more hops a carrier must make and the more the transport pathway resembles a series of resistive steps rather than a smooth highway. In thin stacks this interlayer transport resistance is negligible compared with the barrier-lowering benefit, but as thickness grows the penalty accumulates. Eventually, according to the study’s analysis, the growing interlayer resistance overtakes the diminishing returns of bandgap narrowing, and the total contact resistance begins to rise. The crossover between these two trends lands at three layers, producing the minimum the measurements captured so cleanly.

This competition between bandgap reduction and interlayer transport is more than a curiosity of one material pair; it illustrates a general design principle for two-dimensional electronics. Researchers have long known that layer number tunes the electronic properties of layered semiconductors, but the new results show that the optimal thickness for a device is set not by the channel alone but by the contact physics at its edges. A channel that is ideal for switching characteristics may be far from ideal for injecting current, and vice versa. The authors argue that their findings offer practical guidance for contact engineering and channel thickness selection in next-generation devices, suggesting that designers of MoS2-based transistors should treat trilayer channels with antimony contacts as a particularly favorable combination rather than defaulting to the monolayer devices that dominate much of the literature.

The choice of antimony as the contact metal is itself noteworthy. Metals that form low-barrier contacts to MoS2 without damaging the fragile atomic surface are scarce, and semimetallic electrodes have attracted attention precisely because their modest density of states and weak Fermi-level pinching can keep injection barriers low. The fact that a clear, reproducible layer-dependent signature emerges with antimony electrodes on two different dielectric substrates suggests the effect is a robust feature of the interface, one that other laboratories replicating the fabrication recipe should be able to observe. The transmission line methodology, meanwhile, offers a template for how the field can move beyond anecdotal single-device reports toward systematic, statistically grounded characterizations of contact behavior across thickness series.

For the broader effort to bring two-dimensional semiconductors out of the laboratory and into manufactured electronics, the study lands at an opportune moment. As silicon transistors approach their physical limits, the semiconductor industry has been evaluating atomically thin channels as candidates for future technology nodes, and contact resistance remains one of the most cited roadblocks in those assessments. Results like these sharpen the engineering map: they identify not only which material combinations lower the barrier but also how the internal transport architecture of the channel itself feeds back into contact performance. The three-layer optimum found by Liu and colleagues is unlikely to be the final word, since different contact metals, dielectrics and doping strategies will shift the crossover point, but the underlying principle, that bandgap narrowing and interlayer resistance trade off against each other as thickness grows, gives researchers a quantitative framework for finding the new optimum in each variant. In a field where single atomic layers make the difference between a record-setting device and a mediocre one, knowing exactly where the sweet spot lies is a valuable piece of the puzzle.

Subject of Research: Layer-dependent contact resistance at MoS2–antimony interfaces in two-dimensional field-effect transistors

Article Title: Non-monotonic thickness dependence of MoS 2 –Sb contact resistance driven by bandgap and interlayer transport

Article References: Liu, L., He, Z., Li, Y., Lin, S., & Li, S. (2026). Non-monotonic thickness dependence of MoS2–Sb contact resistance driven by bandgap and interlayer transport. PLOS One, 21(10), e0360960. https://doi.org/10.1371/journal.pone.0360960

Image Credits: AI Generated

DOI: 10.1371/journal.pone.0360960

Keywords: MoS2, contact resistance, two-dimensional semiconductors, Schottky barrier, field-effect transistors, transmission line method, antimony contacts, bandgap engineering, interlayer transport, PLOS One, Non-monotonic, thickness