Polytype-Dependent Upconversion Photoluminescence in 3R-MoS2

Abstract

Ferroelectric van der Waals materials offer switchable polarization states, yet optical readout of their stacking configurations remains challenging. Building on the resonant exciton-exciton annihilation (EEA) mechanism in 2H-phase TMDs, we report the first observation of upconversion photoluminescence (UPL) in rhombohedral MoS2 and demonstrate that this many-body process is strongly polytype-dependent. Using low-temperature spectroscopy, we observe anti-Stokes emission with superlinear power dependence. Beyond serving as a layer-number sensor, UPL provides a sensitive probe of stacking order. Trilayer ABA and BAB polytypes, indistinguishable by surface potential measurements and second harmonic generation, exhibit markedly different UPL intensities, and this persists in thicker samples. First-principles calculations attribute this polytype dependence to modulation of the Gamma-point conduction manifold, which controls energy-matching conditions for the annihilation process. Power-dependent spectroscopy further disentangles two distinct annihilation channels originating from different dark exciton valleys, identified through their contrasting intensity scaling and opposite density-induced energy shifts. Crucially, the annihilation process doubles the energy separation of nearly degenerate dark excitons while converting their weak emission into bright signal, providing experimental access to valley-specific responses that are obscured in direct dark-exciton spectroscopy. Our findings demonstrate that ferroelectric configurations provide a new degree of freedom for controlling nonlinear optical processes, with implications for all-optical ferroelectric readout and electrically switchable wavelength conversion in two-dimensional materials.

Figures (4)

• Figure 1: Polytype-dependent UPL in trilayer 3R-MoS$_2$. (a) Schematic of the four trilayer 3R stacking sequences (ABC, ABA, BAB, CBA), illustrating the relative layer registry and the associated out-of-plane polarization direction (arrows). (b) Optical micrograph image of the trilayer region. (c) KPFM-scan image of the same area used to delineate the domains. (d) Integrated UPL map (2.4-2.8 eV). Color markers indicating the locations assigned to each polytype (Light blue - ABC, dark blue - CBA, red - ABA and Green - BAB). (e) Full PL spectra under 532nm excitation, comparing the four polytypes. The main panel shows the known bright excitons emission ($X_A$, $X_B$), while the insets highlight the momentum-dark exciton band ($X_{\mathrm{Dark}}$) and the upconverted emission band ($X_{\mathrm{UPL}}$).
• Figure 2: Band-structure origin of polytype-dependent UPL in trilayer 3R-MoS$_2$. (a) Schematic of the upconversion mechanism: two momentum-dark excitons formed between the $Q/Q'$ (and/or $K/K'$) conduction valleys and the $\Gamma$ valence band, annihilate into a bright $\Gamma-\Gamma$ exciton, enhanced by the dense cluster of conduction-band states near $\Gamma$. (b) Representative UPL spectra of the neutral configurations (ABA and BAB) with experimental data (symbols) and model curves (solid lines), illustrating the polytype-dependent UPL amplitude and lineshape under identical excitation conditions. (c,d) HSE band structures of trilayer ABA and BAB along $K$--$Q$--$\Gamma$--$Q'$--$K'$. Insets zoom into the conduction-band manifold near $\Gamma$ and mark the expected two-exciton energies for annihilation channels involving $Q-\Gamma$ dark excitons (green dashed line, $2X_{Q\Gamma}$) and $K-\Gamma$ dark excitons (blue dashed line, $2X_{K\Gamma}$). The shaded band around the blue line indicates the assumed broadening used in the resonance model (shown only for $2X_{K\Gamma}$ to avoid visual overload).
• Figure 3: Layer- and power-dependent UPL in 3R-MoS$_2$. (a) Optical micrograph image of the measured flake. (b) False-color UPL-contrast image of the same area, constructed by mapping the spectrally integrated UPL intensity in three energy windows assigned to the R/G/B channels. Integration bands defined in the main text. (c) UPL spectra from $N=3$ and $N=5$--10 layers, showing a pronounced thickness dependence and two spectral components (P1, P2; peak centers marked). Inset: corresponding momentum-dark exciton spectra. (d) Example spectral fits at three excitation power densities for 6 layers (data, total fit, and two components). (e) Log--log excitation power dependence of the integrated intensity, shown separately for P1 (squares) and P2 (triangles) for each layer number, with power-law fits $I\propto P^{\alpha}$ (exponents annotated). (f) Extracted peak energies of P1 (squares) and P2 (triangles) versus excitation power.
• Figure 4: Polytype-dependent UPL in a 7L 3R-MoS$_2$ region. (a) Optical micrograph of the measured 7L area (dashed outlines). (b) KPFM surface-potential map of the same region, with the four measurement locations (D1--D4, circles) used for the optical spectra. (c) UPL spectra acquired at D1--D4, showing strong domain-dependent intensity at fixed thickness. Symbols mark the extracted centers of the two fitted components (P1 - squares, P2 - triangles). (d) KPFM potential profiles along the straight lines connecting the circles in (b). (e) Log--log excitation power dependence of the integrated intensities of P1 (squares) and P2 (triangles) for each domain, with power-law fits overlaid. (f) Extracted peak energies of P1 (squares) and P2 (triangles) versus excitation power for D1--D4.