High harmonic generation has a cutoff law. When an intense laser field ionizes an atom, the freed electron accelerates in the field, reverses direction as the field oscillates, and recombines with the parent ion, emitting a photon whose energy equals the ionization potential plus the kinetic energy gained during the excursion. The maximum kinetic energy — the cutoff — is 3.17 times the ponderomotive energy, which scales as the intensity times the square of the wavelength. This three-step model, established in the 1990s, sets a ceiling on the photon energies that high harmonic generation can produce for a given laser. Higher energy requires longer wavelength or higher intensity. The scaling is universal for atoms.
Dubey and Neufeld (arXiv 2602.22599, February 2026) find that fullerenes break through this ceiling through a mechanism that inverts the standard scaling.
Gas-phase fullerenes from C20 to C60, driven by near-infrared laser pulses at 10^14 W/cm2, produce a standard harmonic plateau followed by an anomalous second plateau extending to 115 electron volts — well above the three-step cutoff. The second plateau has four properties that collectively rule out any explanation within the standard model.
First, it operates through a recombination mechanism, not through the rescattering process that drives normal high harmonics. Second, the cutoff scales inversely with wavelength: shorter wavelengths produce higher cutoff energies, the opposite of the standard law where longer wavelengths drive higher harmonics. Third, the cutoff depends linearly on the electric field amplitude rather than quadratically through the ponderomotive energy. Fourth, the emission cannot be explained by classical real-space electron trajectories — the standard picture of the electron leaving, gaining kinetic energy, and returning.
The source of the anomalous plateau is quantum mechanical resonances specific to the fullerene cage structure. The carbon cage supports sharp electronic resonances at specific energies, determined by the molecular geometry rather than by the laser parameters. When the laser drives electron dynamics within the fullerene, these resonances enable coherent emission at energies that the three-step model does not predict, because the emission pathway does not involve the classical excursion of a free electron in the laser field.
The inverted wavelength scaling is the key signature. In the three-step model, longer wavelength means more time for the electron to accelerate, means more kinetic energy, means higher cutoff. In the resonance mechanism, longer wavelength means the laser frequency is further from the molecular resonance, reducing the efficiency of the resonant emission channel. The two mechanisms compete: the standard plateau extends with wavelength, the anomalous plateau recedes. At visible wavelengths, the anomalous plateau dominates. At mid-infrared wavelengths, the standard plateau catches up.
The practical implication is coherent broadband extreme ultraviolet emission from a near-infrared driver without the usual penalty of requiring mid-infrared or X-ray wavelengths. The fullerene's molecular structure does the work that would otherwise require a more powerful or more exotic laser. The ceiling is not gone — but for this class of targets, the rules for where it sits are different.