Atomic-scale sensing of photoexcitation processes in electronically isolated molecules via atomic force microscopy

Technical Summary: Atomic-scale sensing of photoexcitation processes in electronically isolated molecules via atomic force microscopy

Problem Statement
While scanning tunnelling microscopy (STM) combined with optical techniques has provided atomic-scale insights into photophysical processes, a significant gap remains in the ability to study photo-induced electron transfer in fully electronically isolated molecules. Previous STM-based approaches typically require molecules to be decoupled from metal substrates using thin insulating spacers (e.g., 3–4 monolayers of NaCl). However, the finite conductance required for STM operation limits the lifetime of photoexcited states due to charge exchange with the substrate, often restricting lifetimes to the picosecond or low nanosecond range. Furthermore, existing methods struggle to access regimes where molecules are completely isolated from the substrate, which is crucial for understanding charge separation and photocatalytic processes without substrate interference.

Methodology
The authors introduce Photoexcitation Single-Charge Atomic Force Microscopy (PE-AFM) to detect photoexcitation in single molecules adsorbed on thick, non-conductive dielectric films (specifically >14 monolayers of NaCl on Cu(111)). The methodology involves:

• Synchronization: Nanosecond laser pulses (centered at 640 nm, 1.94 eV) are synchronized with the oscillation of the AFM cantilever. Pulses illuminate the tip-sample junction only at the cantilever's zero crossings.
• Charge State Modulation: Photoexcitation of copper phthalocyanine (CuPc) molecules drives electron transfer between the tip and the molecule, changing the molecular charge state. A gate voltage ($V_G$) is applied to tune the energetic alignment of states, steering the subsequent neutralization (discharging) process to occur predominantly when the tip is near the bottom turnaround point of its oscillation.
• Signal Detection: The asymmetry in the molecular charge state between the inward and outward motion of the cantilever results in a damping of the cantilever's motion. This damping is detected as an increase in the driving amplitude ($A_{drive}$) required to maintain oscillation.
• Simulation and Spectroscopy: Finite-element simulations (COMSOL) were used to model near-field enhancement. Additionally, pump-probe voltage-pulse sequences were employed to extract the lifetimes and energies of electronic excited states, and a many-body theoretical description was used to rationalize the observed pathways.

Key Results

• Near-Field Enhancement: Simulations revealed that while metallic substrates screen in-plane electric fields ($E_{\parallel}$), thick NaCl films allow partial penetration of these fields. Peak near-field enhancement ratios reached ~1200 for the tip-NaCl-Cu geometry, sufficient to drive photoexcitation despite the electronic isolation.
• Polarization Dependence: The PE-AFM signal exhibited strong dependence on light polarization, maximizing for p-polarized light and vanishing for s-polarized light. This confirmed the signal origin as molecular photoexcitation rather than thermal expansion or direct cantilever excitation.
• Sub-molecular Imaging: PE-AFM imaging of CuPc revealed a ring-like contrast at the molecular periphery at certain gate voltages, evolving to a central signal at higher gate voltages. Unlike STM-based photocurrent images which often resolve orbital nodal planes, the PE-AFM contrast was "smeared" and lacked well-defined nodal structures.
• Long-Lived State Identification: The authors identified that the observed contrast arises from a long-lived excited state. Pump-probe spectroscopy measured the lifetime of the quartet state ($Q_1$) to be $11 \pm 3$ $\mu$s. This lifetime is orders of magnitude longer than those typically observed in STM experiments on thinner films, attributed to the suppression of substrate-induced quenching.
• Mechanism of Contrast: The "smeared" sub-molecular contrast is attributed to signal saturation. Because the $Q_1$ state is long-lived, the molecule remains in a charged state for a significant fraction of the cantilever period regardless of the exact tip position during the initial photoexcitation. Consequently, the spatial contrast is dominated by the tip-position-dependent discharging rate (tunnelling) rather than the photoexcitation probability itself.
• Pathway Rationalization: The data supports a pathway where photoexcitation populates a doublet state ($D_2$), which undergoes internal conversion and intersystem crossing to the long-lived quartet state ($Q_1$). The subsequent tunnelling event from $Q_1$ to the ground state ($D_0$) is the rate-limiting step that generates the AFM signal.

Significance and Claims
The paper claims to establish a powerful platform for investigating photophysical processes at the single-molecule level, specifically for electronically isolated systems. The primary significance lies in:

1. Accessing Isolated Regimes: PE-AFM enables the study of photo-induced charge transfer in molecules that are fully decoupled from metallic substrates, preventing exciton quenching and allowing for the observation of intrinsic long-lived excited states (such as the $Q_1$ state with a ~11 $\mu$s lifetime).
2. Comprehensive Characterization: By combining PE-AFM with excited-state spectroscopy and lifetime measurements, the technique allows for the experimental identification of photoexcitation pathways, including the detection of dark states and the characterization of intersystem crossing.
3. Sub-molecular Resolution: The technique achieves atomic-scale spatial resolution (ångström scale) for photoexcitation processes, demonstrating that single-electron photocurrents can be detected via AFM on insulating surfaces.
4. Mechanistic Insight: The work provides a clear explanation for the nature of the observed contrast, linking the "smeared" orbital features to the saturation effects caused by long-lived states, distinguishing it from previous STM-based observations where shorter lifetimes allowed for orbital-resolved imaging.

The authors conclude that this approach opens the door to systematically investigating how different charge configurations and molecular environments influence photoexcitation pathways and relaxation dynamics, with potential extensions to studying charge and energy transfer in molecular assemblies.