Endothermic Charge Separation Occurs Spontaneously in Non-Fullerene Acceptor/Polymer Bulk Heterojunction.

Organic photovoltaics (OPVs) based on non-fullerene acceptors (NFAs) have achieved a power conversion efficiency close to 20%. These NFA OPVs can generate free carriers efficiently despite a very small energy level offset at the donor/acceptor interface. Why these NFAs can enable efficient charge separation (CS) with low energy losses remains an open question. Here, the CS process in the PM6:Y6 bulk heterojunction is probed by time-resolved two-photon photoemission spectroscopy. It is found that the CS, the conversion from bound charge transfer (CT) excitons to free carriers, is an endothermic process with an enthalpy barrier of 0.15 eV. The CS can occur spontaneously despite being an endothermic process, which implies that it is driven by entropy. It is further argued that the morphology of the PM6:Y6 film and the anisotropic electron delocalization restrict the electron and hole wavefunctions within the CT exciton such that they can primarily contact each other through point-like junctions. This configuration can maximize the entropic driving force.

Enthalpy-uphill exciton dissociation in organic/2D heterostructures promotes free carrier generation.

Despite the large binding energy of charge transfer (CT) excitons in type-II organic/2D heterostructures, it has been demonstrated that free carriers can be generated from CT excitons with a long lifetime. Using a model fluorinated zine phthalocyanine (F8ZnPc)/monolayer-WS2 interface, we find that CT excitons can dissociate spontaneously into free carriers despite it being an enthalpy-uphill process. Specifically, it is observed that CT excitons can gain an energy of 250 meV in 50 ps and dissociate into free carriers without any applied electric field. This observation is surprising because excited electrons typically lose energy to the environment and relax to lower energy states. We hypothesize that this abnormal enthalpy-uphill CT exciton dissociation process is driven by entropy gain. Kinetically, the entropic driving force can also reduce the rate for the reverse process - the conversion of free electron-hole pairs back to CT excitons. Hence, this mechanism can potentially explain the very long carrier lifetime observed in organic/2D heterostructures.

Nanoscale Periodic Trapping Sites for Interlayer Excitons Built by Deformable Molecular Crystal on 2D Crystal.

The nanoscale moiré pattern formed at 2D transition-metal dichalcogenide crystal (TMDC) heterostructures provides periodic trapping sites for excitons, which is essential for realizing various exotic phases such as artificial exciton lattices, Bose-Einstein condensates, and exciton insulators. At organic molecule/TMDC heterostructures, similar periodic potentials can be formed via other degrees of freedom. Here, we utilize the structure deformability of a 2D molecular crystal as a degree of freedom to create a periodic nanoscale potential that can trap interlayer excitons (IXs). Specifically, two semiconducting molecules, PTCDI and PTCDA, which possess similar band gaps and ionization potentials but form different lattice structures on MoS2, are investigated. The PTCDI lattice on MoS2 is distorted geometrically, which lifts the degeneracy of the two molecules within the crystal's unit cell. The degeneracy lifting results in a spatial variation of the molecular orbital energy, with an amplitude and periodicity of ∼0.2 eV and ∼2 nm, respectively. On the other hand, no such energy variation is observed in PTCDA/MoS2, where the PTCDA lattice is much less distorted. The periodic variation in molecular orbital energies provides effective trapping sites for IXs. For IXs formed at PTCDI/MoS2, rapid spatial localization of the electron in the organic layer toward the interface is observed, which demonstrates the effectiveness of these interfacial IX traps.

Hybrid Heterostructures to Generate Long-Lived and Mobile Photocarriers in Graphene.

We report the generation of long-lived and highly mobile photocarriers in hybrid van der Waals heterostructures that are formed by monolayer graphene, few-layer transition metal dichalcogenides, and the organic semiconductor F8ZnPc. Samples are fabricated by dry transfer of mechanically exfoliated MoS2 or WS2 few-layer flakes on a graphene film, followed by deposition of F8ZnPc. Transient absorption microscopy measurements are performed to study the photocarrier dynamics. In heterostructures of F8ZnPc/few-layer-MoS2/graphene, electrons excited in F8ZnPc can transfer to graphene and thus be separated from the holes that reside in F8ZnPc. By increasing the thickness of MoS2, these electrons acquire long recombination lifetimes of over 100 ps and a high mobility of 2800 cm2 V-1 s-1. Graphene doping with mobile holes is also demonstrated with WS2 as the middle layers. These artificial heterostructures can improve the performance of graphene-based optoelectronic devices.

Defect-Polaron and Enormous Light-Induced Fermi-Level Shift at Halide Perovskite Surface.

Halide perovskites intrinsically contain a large amount of point defects. The interaction of these defects with photocarriers, photons, and lattice distortion remains a complex and unresolved issue. We found that for halide perovskite films with excess halide vacancies, the Fermi level can be shifted by as much as 0.7 eV upon light illumination. These defects can trap photocarriers for hours after the light illumination is turned off. The enormous light-induced Fermi level shift and the prolonged electron trapping are explained by the capturing of photocarriers by halide vacancies at the surface of the perovskite film. The formation of this defect-photocarrier complex can result in lattice deformation and an energy shift in the defect state. The whole process is akin to polaron formation at a defect site. Our data also suggest that these trapped carriers increase the electrical polarizability of the lattice, presumably by enhancing the defect migration rate.

Extracting Electrons from Delocalized Excitons by Flattening the Energetic Pathway for Charge Separation.

At organic donor-acceptor (D-A) interfaces, electron and hole are bound together to form charge transfer (CT) excitons. The electron and hole wave functions in these CT excitons can spatially delocalize. The electron delocalization opens up possibilities of extracting free charges from bound excitons by manipulating the potential energy landscape on the nanoscale. Using a prototype trilayer structure that has a cascade band structure, we show that the yield of charge separation can be doubled as compared to the bilayer counterpart when the thickness of the intermediate layer is around 3 nm. This thickness coincides with the electron delocalization size of CT excitons typically found in these organic films. Tight-binding calculation for the CT states in the trilayer structure further demonstrates that electron delocalization, together with the energy level cascade, can effectively flatten the energetic pathway for charge separation. Hence, it is possible to add nanometer-thick layers between the donor and the acceptor to significantly enhance the charge separation yield.

Collective Effects of Band Offset and Wave Function Dimensionality on Impeding Electron Transfer from 2D to Organic Crystals.

Excited-state electron transfer (ET) across molecules/transition metal dichalcogenide crystal (TMDC) interfaces is a critical process for the functioning of various organic/TMDC hybrid optoelectronic devices. Therefore, it is important to understand the fundamental factors that can facilitate or limit the ET rate. Here it is found that an undesirable combination of the interfacial band offset and the spatial dimensionality of the delocalized electron wave function can significantly slow down the ET process. Specifically, it is found that whereas the ET rate from TMDCs (MoS2 and WSe2) to fullerenes is relative insensitive to the band offset, the ET rate from TMDCs to perylene molecules can be reduced by an order of magnitude when the band offset is large. For the perylene crystal, the sensitivity of the ET rate on the band offset is explained by the 1D nature of the electronic wave function, which limits the availability of states with the appropriate energy to accept the electron.