Controlled single-electron transfer enables time-resolved excited-state spectroscopy of individual molecules

Scanning-probe microscopy combined with spectroscopic techniques offers atomic-scale insights into single-molecule properties. However, unambiguously assigning observed signals to specific electronic quantum transitions remains a challenge. For instance, the interpretation of scanning tunneling microscopy (STM) luminescence signals in PTCDA as either phosphorescence or trion-related fluorescence is debated in the literature. Similarly, in scanning tunneling spectroscopy (STS), distinguishing higher-lying excited states from multiple charging effects is difficult because the molecule usually reverts to its ground state quickly after a tunneling event. The charge exchange with the surface is often elusive, making it hard to differentiate processes. For example, a feature at a higher bias voltage than the LUMO could be either LUMO+1 or a second electron in the LUMO from a transient charge state. Even combining luminescence with tunneling spectroscopy doesn't resolve individual electronic transitions because multiple transitions contribute to a single spectral feature. This research aims to overcome these limitations by developing a single-molecule spectroscopy method that allows the individual study of many quantum transitions of different types, providing a more complete and accurate understanding of single-molecule electronic properties.

Existing single-molecule studies using scanning-probe microscopy have achieved significant progress in structure determination, orbital density imaging, and electron-spin resonance. Integration of optical spectroscopy (Raman and luminescence) into local-probe microscopy has advanced atomic-scale insights into light-matter interactions, allowing for the direct visualization and testing of theoretical concepts regarding light emission from organic materials. However, the unambiguous assignment of experimental observations to specific electronic quantum transitions remains a significant challenge. Contradictory interpretations exist regarding the origin of STM luminescence signals in PTCDA, with some attributing it to phosphorescence and others to trion-related fluorescence. Similarly, in STS, differentiating higher-lying excited states from multiple charging effects is often difficult. The use of thick insulating films to prevent tunneling to the substrate allows for the separation of individual electronic transitions by controlling single-electron tunneling between an AFM tip and the molecule, making different charge states accessible. However, mapping the excitation spectrum of an individual adsorbed molecule remained elusive until the present work.

This study introduces a single-molecule spectroscopy method based on controlled charge exchange between an AFM tip and a molecule adsorbed on a thick NaCl film (electrically isolating the molecule from the substrate). The technique involves a pump-probe voltage pulse sequence: a set pulse brings the molecule to a defined state, a sweep pulse with variable voltage and time excites the molecule, and a read-out voltage pulse maps the resulting charge state onto two distinguishable states via AFM force detection. The read-out voltage is tuned to the charge-degeneracy point between the neutral and a singly charged state (e.g., S0 and D0 or D0-). The large electron-phonon coupling prevents interconversion between degenerate charge states by tunneling, leading to a Gaussian-shaped transition probability. The scheme also differentiates states with the same charge, like S0 and T1, by projecting them onto different charge states. A many-body description of the electronic states is employed to interpret the results. The applied gate voltage shifts the molecular energy levels with respect to the tip's Fermi level, changing the energy of many-body states with different net charges. The lever arm (α), representing the fraction of gate voltage across the NaCl film, is determined experimentally. By repeatedly applying the pulse sequence with varying sweep voltages and times, the relative populations of the different charge states during the readout phase are measured as a function of the sweep voltage. These measurements are then fitted to a set of differential rate equations that model the possible electronic transitions, yielding the energy-level alignment of the low-lying electronic states. The lever arm is calibrated using previously determined S0-S1 energy differences from STM-induced luminescence experiments. The measurements are repeated on several molecules to account for potential environmental influences.

The method was successfully applied to pentacene and PTCDA. For pentacene, the experiments, initializing in both cationic (D0+) and neutral states (partially populating S0 and T1), yielded a many-body energy diagram, revealing the relative energies of several states, including the S0, T1, D0, D0+, and D1 states. The S0-T1 energy difference (0.90 ± 0.06 eV) and the D0-D1 energy difference (0.99 ± 0.04 eV) were determined. The reorganization energies for several redox transitions were also extracted. For PTCDA, the experiments, initializing in both cationic (D0+) and neutral (partially populated S0 and T1) states provided a many-body energy diagram. The S0-S1, S0-T1, and D0-D1 energy differences were determined as (2.39 ± 0.11) eV, (1.28 ± 0.07) eV, and (1.34 ± 0.08) eV respectively. These results align closely with STM-induced luminescence data, suggesting that both T1→S0 (phosphorescence) and D1→D0 (fluorescence of the anion) transitions could contribute to the 1.33 eV luminescence signal. The lifetime of the T1 state was measured to be approximately 300 µs on >20 ML NaCl, contrasting with shorter lifetimes (100 ps) observed on thinner NaCl films, providing insights into the dynamics of luminescence pathways. The study further extrapolates the energy level alignment to ultrathin NaCl films to explain the observed STM-induced luminescence in such systems, showing two possible pathways (via S0 or D1) that could lead to the 1.33 eV luminescence signal.

The presented single-molecule spectroscopy method successfully addresses the limitations of previous techniques by allowing the individual characterization of various electronic transitions within a single molecule. The ability to extract energy levels, reorganization energies, and relative rates of competing transitions provides a detailed understanding of the molecule's electronic structure and dynamics. The agreement between experimental energy differences and previously reported STM-induced luminescence data validates the method's accuracy and resolves the ambiguity in interpreting the 1.33 eV luminescence signal in PTCDA. The differences in T1 state lifetimes on thick and thin NaCl films highlights the role of the substrate in influencing the luminescence pathways. The ability to prepare and study molecules in specific excited states is a significant advance that will greatly benefit future studies on tip-induced chemical reactions and molecular luminescence.

This work introduces a novel single-molecule spectroscopy method capable of resolving individual electronic transitions, including radiative, non-radiative, and redox processes. Its application to pentacene and PTCDA elucidates the complexities of excited-state dynamics and resolves ambiguities in the interpretation of previous STM luminescence experiments. The method's ability to control and study molecules in specific excited states opens up new possibilities for studying and engineering tip-induced chemical reactions and molecular luminescence. Future work could involve extending this approach to a wider range of molecules and to the study of multiple-charged states, offering richer insights into the quantum mechanical behavior of individual molecules.

While the study provides detailed insights, some limitations exist. The analysis relies on a model based on a set of differential rate equations, and the accuracy of the derived energy levels and rates depends on the validity of these equations. Potential interactions between the molecule and the AFM tip, such as Stark shifts or charge transfer effects, could influence the measurements, though efforts were made to minimize these effects by careful tip preparation and pulse sequence design. The extrapolation to ultrathin films is based on assumptions and might not fully capture the complex interplay of factors influencing the luminescence signal in those systems. The study focuses on two model systems; broader application to a wider range of molecules needs further investigation.