Tunable electronic energy level alignment and exciton diversity in organic-inorganic van der Waals heterostructures

TL;DR

The paper addresses how to engineer electronic energy level alignment and exciton diversity in 2D organic–inorganic van der Waals heterostructures. It employs ab initio GW-BSE calculations on four bilayers (PTCDA or PDI on MoS2 or WS2) with an efficient dielectric environment treatment and exciton-decomposition analysis to elucidate interfacial effects. Key findings include up to ~1 eV renormalization of molecular gaps due to nonlocal screening, a switch from type-I to type-II ELAs depending on the TMD, and a rich excitonic landscape comprising intralayer, interlayer, and hybrid excitons; notably, the PDI–WS2 interlayer exciton exhibits high binding energy, small real-space extent, long radiative lifetime, and strong polarization anisotropy. These results establish organic–inorganic bilayers as tunable platforms for exciton-based quantum and optoelectronic devices, with potential for exciton condensation and efficient charge separation in photovoltaic applications.

Abstract

van der Waals stacking of two-dimensional (2D) materials offers a powerful platform for engineering material interfaces with tailored electronic and optical properties. While most van der Waals multilayers have featured inorganic monolayers, incorporating molecular monolayers introduces new degrees of tunability and functionality. Here, we investigate hybrid bilayers composed of atomically thin perylene-based molecular crystals interfaced with monolayer transition metal dichalcogenides (TMDs), specifically MoS2 and WS2. Using ab initio many-body perturbation theory within the GW approximation and the Bethe-Salpeter equation approach, we predict emergent properties beyond those of the isolated constituent systems. Notably, we find substantial renormalization of monolayer molecular crystal band gap due to TMD-induced polarization. Furthermore, by varying the TMD monolayer, we demonstrate tuning of the energy level alignment of the bilayer and subsequent control over a diversity of lowest-energy excitons, which include strongly bound hybrid excitons and long-lived charge-transfer excitons. These findings establish organic-inorganic van der Waals heterostructures as a promising class of materials for tunable optoelectronic devices and quantum excitonic phenomena, expanding the design space for low-dimensional systems.

Figures (5)

• Figure 1: Room-temperature STM image of (a) PTCDA and (b) PDI, grown on monolayer MoS$_2$ on a SiO$_2$ substrate, adapted from Ref. chowdhury2024. Schematic atomic structure of (c) PTCDA $-$ with herringbone 2D arrangement,(d) PDI $-$ with brick-wall 2D arrangement, and (e) PDI-WS$_2$ interface supercell. Structural parameters are indicated with blue labels.
• Figure 2: Computed GW band structure for (a) PDI-MoS$_2$, (c) PTCDA-MoS$_2$, (e) PDI-WS$_2$, and (h) PTCDA-WS$_2$. Band colors result from the projection of interfacial orbitals onto orbitals of the individual layers. Schematic of band edges for (b) PDI-MoS$_2$, (d) PTCDA-MoS$_2$, (f) PDI-WS$_2$, and (i) PTCDA-WS$_2$. DFT-calculated real-space squared wave functions of the VBM-HOMO at K point (shown by yellow circle) for (g) PDI-WS$_2$ and (j) PTCDA-WS$_2$$-$ top and side views. Spin-orbit coupling effects are not included.
• Figure 3: For (a) PDI-MoS$_2$ and (b) PDI-WS$_2$ interfaces, the imaginary part of the dielectric function, $\epsilon_2$, is plotted as a function of the photon energy, showing the contributions from four distinct excitonic species, schematically represented in (c): TMD intralayer (blue), molecular intralayer (red), low-energy interlayer (green), and high-energy interlayer (magenta) excitons. The dominant nature of each peak is indicated on the figure with "D" for direct (intralayer), "H" for hybrid, and "I" for interlayer excitons.
• Figure 4: For the PDI-WS$_2$ bilayer, we show the isosurface maps of the exciton wavefunctions (selected to include 98% of the electron density), showing the electron probability density for the most probable hole positions: (a) I1 interlayer, (b) D2 WS$_2$ intralayer, and (c) H3 hybrid excitons. A $20 \times 10 \times 1$ supercell is used to generate these isosurface plots. The exciton Bohr radii (quantitatively defined in SI) are shown with the gray arrows. (d) Electron-hole correlation function along the $z$-axis. (e) Distribution of exciton radiative lifetimes with respect to the photon energy, with the intralayer, interlayer, and hybrid excited states shown in blue, green, and red, respectively.
• Figure 5: For (a) PDI-MoS$_2$ and (b) PDI-WS$_2$ systems, we show the GW-BSE computed imaginary part of the dielectric function, $\epsilon_2$, as a function of the photon energy, for two orthogonal polarization directions (see inset). The prominent features are indicated by vertical dashed lines, with "D" for direct (intralayer), "H" for hybrid, and "I" for interlayer excitons.