Atomically Thin <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML" display="inline"> <mml:msub> <mml:mi>MoS</mml:mi> <mml:mn>2</mml:mn> </mml:msub> </mml:math> : A New Direct-Gap Semiconductor

This Physical Review Letters article asks a central question in two-dimensional (2D) materials science: how do the electronic and optical band structures of molybdenum disulfide (MoS2) evolve as the material is thinned from bulk-like crystals down to a single SMoS monolayer? The question matters because the optical usefulness of a semiconductor depends strongly on whether its lowest-energy electronic transition is direct (efficient light emission/absorption) or indirect (phonon-assisted and typically weak). In bulk MoS2, the material is an indirect-gap semiconductor with a band gap of 1.29 eV, which limits photoluminescence (PL) efficiency. If quantum confinement in atomically thin layers can drive an indirect-to-direct band-gap crossover, MoS2 could become a bright, room-temperature light emitter and a more practical platform for optoelectronics and photonics.

The authors study ultrathin MoS2 with thicknesses corresponding to the number of SMoS layers, denoted by $N=1,2,\dots ,6$. They emphasize that samples are prepared both on solid surfaces and as free-standing films, aiming to minimize substrate-induced perturbations. The methodology is experimental and triangulates the electronic structure using three complementary optical spectroscopies: absorption spectroscopy, photoluminescence (PL), and photoconductivity spectroscopy. Thickness is characterized by atomic-force microscopy (AFM), and optical contrast is also used to determine monolayer count with monolayer accuracy. For PL quantum yield (QY), suspended samples are excited with a continuous-wave 532 nm laser at low power (about 50 W on the sample) to avoid heating and saturation. QY is calibrated using a rhodamine 6G thin film standard whose QY is independently related to that of a dilute rhodamine 6G solution (assumed close to 1). Photoconductivity is measured using two-terminal devices fabricated from mono- and bilayer MoS2 on oxide-covered Si, with monochromatic, modulated laser excitation and lock-in detection.

The key findings are (i) a strong thickness-dependent upward shift of the indirect band gap and (ii) a comparatively small change in the direct band gap, producing an indirect-to-direct crossover at the monolayer limit. Specifically, the indirect gap shifts upward from the bulk value of 1.29 eV to 1.90 eV as $N$ decreases, i.e., an increase of more than 0.6 eV. Over the same thickness range, the direct gap increases by only about 0.1 eV. Because the indirect gap rises much more than the direct gap, the lowest-energy transition becomes direct in the monolayer.

Optically, the PL and absorption spectra reveal this crossover. For suspended monolayers, the PL spectrum contains a single narrow feature centered at 1.90 eV with a linewidth of about 50 meV. This PL peak matches the lower-energy direct absorption resonance in both position and width, so the monolayer PL is attributed to direct-gap luminescence. For few-layer samples, the PL becomes multi-peaked: peak A aligns with the monolayer emission peak and redshifts/broadens slightly with increasing $N$; peak B lies about 150 meV above peak A; and a broad feature I lies below peak A and shifts systematically to lower energies as $N$ increases. The position of feature I for $N=2$ to 6 approaches the bulk indirect gap of 1.29 eV, supporting its assignment to indirect-gap luminescence.

The most striking optical consequence is the dramatic enhancement of PL quantum efficiency. The authors report that the PL QY for suspended samples decreases steadily with increasing thickness from $N=1$ to $N=6$. They estimate that for $N=2$ to 6, the PL QY is on the order of $10^{-5}$ to $10^{-6}$, while for the monolayer they observe a QY as high as $4\times 10^{-3}$. This corresponds to an increase by more than a factor of 1000 when comparing the monolayer to the bulk crystal (which is described as a dark, indirect-gap emitter with negligible QY). The paper also notes that PL spectra for monolayer and few-layer samples are distinct, consistent with a change in the dominant recombination pathway.

To directly probe the indirect transition energies, the authors use photoconductivity spectroscopy because absorption sensitivity is limited for indirect transitions. Their photoconductivity spectra show that for bilayer MoS2, the onset of photoconductivity occurs around 1.60 eV, coinciding with PL peak I, and the conductivity increases slowly with photon energy toward the direct gap, a behavior characteristic of indirect-gap materials. In contrast, for monolayer MoS2, well below the direct-gap transition there is no appreciable photoconductive response; instead, the photoconductivity exhibits an abrupt increase only near the direct band gap. This spectral contrast leads to the conclusion that bilayer MoS2 remains indirect-gap, whereas monolayer MoS2 behaves as a direct-gap semiconductor.

The authors support this conclusion with a simplified model for 2D semiconductor absorbance: direct transitions are treated as step-function-like in energy (proportional to a joint density of states factor), while indirect transitions are treated as phonon-assisted processes with temperature-broadened onset. They state that their model (neglecting excitonic effects and assuming constant matrix elements) describes the experimental photoconductivity spectra well: for bilayer, both indirect and direct contributions are required, whereas for monolayer, a direct-gap transition alone suffices. They also discuss the indirect-to-direct crossover as a consequence of quantum confinement and a zone-folding scheme applied to the bulk band structure. In this picture, monolayer thickness corresponds to allowed quantized momentum cuts that pass through the H and A points, placing both the conduction band minimum and valence band maximum at the H point, yielding a direct gap. As $N$ increases, the allowed momentum cuts shift toward the bulk-like $\Gamma$-to-K direction, restoring an indirect gap.

Limitations are not presented as a formal list, but they are implicit in the methodology and modeling. The photoconductivity analysis explicitly neglects excitonic effects and the variation of optical matrix elements with energy, which could affect quantitative lineshapes and onset energies. The paper also relies on room-temperature optical measurements, where phonon assistance and thermal broadening can blur distinctions between transitions. Additionally, the study focuses on $N\le 6$ and on mechanically exfoliated samples; while this is appropriate for demonstrating the crossover, it may limit generality regarding strain, defects, and substrate effects, even though the authors include free-standing measurements to reduce substrate perturbations.

Practically, the results are important because they establish monolayer MoS2 as an efficient, room-temperature light emitter, unlike bulk MoS2. This makes it a promising platform for photostable markers and sensors, and it suggests routes to optimize band gaps for photocatalysis and photovoltaic applications. More broadly, the paper argues that the distinctive electronic properties of atomically thin materials are not unique to graphene; they extend to other van der Waals-bonded solids, expanding the design space for 2D optoelectronic materials.

Overall, the paper provides a clear experimental demonstration that quantum confinement in MoS2 drives an indirect-to-direct band-gap crossover at the monolayer, accompanied by a dramatic ( orders of magnitude) enhancement in PL quantum yield and a direct optical signature in both emission and photoconductivity spectra.