Asymmetric laser pulses let scientists read molecular vibrations buried in metallic noise

Research by the Institute for Molecular Science and Tohoku University. Published February 19, 2026 in The Journal of Chemical Physics as a Featured Article.

The signal drowned by its own substrate

Surface scientists have long faced an irritating paradox. The technique best suited for reading molecular vibrations at surfaces - vibrational sum frequency generation (SFG) spectroscopy - works beautifully on dielectric substrates but becomes nearly useless on metal ones. The problem is that metals generate a massive non-resonant background signal from their free electrons, and this background swamps the faint vibrational signatures of whatever molecules sit on top. Since many of the surfaces that matter most in catalysis and materials science are metallic, this limitation has been a persistent frustration.

A team from Japan's Institute for Molecular Science and Tohoku University has now found an elegant workaround. By reshaping one of the laser pulses used in the experiment - giving it a steep rise and a slow, gradual decay - they exploited a timing difference between the metal's response and the molecules' response. The metal background vanishes almost instantaneously. Molecular vibrations linger. Wait a carefully controlled interval, and the molecules are the only things still talking.

How timing separates molecule from metal

Sum frequency generation spectroscopy works by overlapping two laser pulses at a surface: one in the visible or near-infrared range and one in the infrared. When they arrive at the surface together, they generate a signal at the sum of their frequencies - but only from molecules at the interface, not from the bulk material. This surface specificity is what makes SFG powerful.

The Japanese team combined SFG with a scanning tunneling microscope (STM) to achieve nanoscale spatial resolution - a technique called tip-enhanced SFG (TE-SFG). The sharp metallic tip of the STM concentrates the electromagnetic field into a nanoscale gap between tip and substrate, amplifying the signal from whatever molecules live in that gap. The enhancement factor, estimated from the experimental data, is on the order of 10 million.

But that same metallic tip and metallic substrate also generate the overwhelming non-resonant background. Here is where the pulse shaping becomes critical. The researchers crafted their near-infrared pulse into what they call a temporally asymmetric waveform: it rises sharply and decays slowly. They then introduced a controlled time delay between this pulse and the infrared pulse.

The physics is straightforward. The metal's free-electron response tracks the laser field almost instantaneously - it rises and falls with the pulse. Molecular vibrations, by contrast, are resonant phenomena that ring like a struck bell, persisting well after the driving field has passed. By delaying the infrared pulse so that it arrives after the near-infrared pulse's steep rise has passed but while the molecular vibrations still persist, the researchers capture molecular signal while the metal background has already died away.

Aromatic ring vibrations emerge from the noise

The payoff was immediate and concrete. Using this approach on molecular samples adsorbed on metallic surfaces, the team detected weak vibrational modes originating from aromatic rings - signals that their previous TE-SFG experiments had failed to capture. These are not exotic laboratory curiosities. Aromatic compounds are ubiquitous in catalysis, organic electronics, and biological surface chemistry. Being able to read their vibrational signatures at the nanoscale, on metal substrates, opens analytical possibilities that did not previously exist.

The team also demonstrated something that standard SFG cannot easily achieve: determining the absolute orientation of molecules relative to the surface. By analyzing the interference pattern between the molecular vibrational signal and the residual background, they could distinguish whether a molecule points "up" or "down" - not just its tilt angle, but its actual direction. This is valuable information for understanding how molecules anchor to catalytic surfaces, how self-assembled monolayers organize, and how molecular devices function.

Confirming the signal comes from the nanoscale gap

One persistent concern with tip-enhanced spectroscopy techniques is whether the detected signal genuinely originates from the nanoscale region between tip and substrate, or whether it includes far-field contributions from a larger area. The researchers addressed this by simultaneously detecting signals scattered in both forward and backward directions. Both signals matched in their spectral features, confirming that they arose from the near-field enhancement in the tip-substrate gap rather than from far-field illumination of a broader surface area.

This verification is important because the spatial resolution claims of tip-enhanced techniques rest on the assumption that the signal comes exclusively from the enhanced region. If far-field contributions leak in, the effective resolution degrades to the diffraction limit - micrometers rather than nanometers.

The road to watching reactions in real time

The researchers outline an ambitious next step: sweeping the time delay between pulses to perform time-resolved measurements. In principle, this would allow them to track ultrafast molecular dynamics and chemical reactions at surfaces with both nanoscale spatial resolution and femtosecond temporal resolution. That combination - seeing where molecules are and what they are doing, at the nanometer and femtosecond scale - would be a powerful tool for studying catalytic mechanisms.

But important limitations remain. The current work was performed under ultrahigh vacuum conditions with well-defined model systems. Extending the technique to ambient conditions, liquid environments, or operando catalysis measurements would introduce substantial technical challenges. The signal levels, while enhanced ten-million-fold, still require long acquisition times for weak vibrational modes. And the STM-based geometry restricts measurements to conductive substrates and limits the types of samples that can be studied.

The technique also requires sophisticated pulse-shaping optics and precise timing control, which limits its accessibility to laboratories with substantial ultrafast laser infrastructure. It is not yet a routine analytical tool. Still, for the subset of surface science questions where knowing the exact orientation and vibrational character of individual molecules at nanoscale precision matters, this approach addresses a real and previously unsolved measurement problem.