Atomically precise electronics operating at optical frequencies require tools that can characterize them on their intrinsic length and time scales to guide device design. Lightwave-driven scanning tunnelling microscopy is a promising technique towards this purpose. It achieves simultaneous sub-ångström and sub-picosecond spatio-temporal resolution through ultrafast coherent control by single-cycle field transients that are coupled to the scanning probe tip from free space. Here, we utilize lightwave-driven terahertz scanning tunnelling microscopy and spectroscopy to investigate atomically precise seven-atom-wide armchair graphene nanoribbons on a gold surface at ultralow tip heights, unveiling highly localized wavefunctions that are inaccessible by conventional scanning tunnelling microscopy. Tomographic imaging of their electron densities reveals vertical decays that depend sensitively on wavefunction and lateral position. Lightwave-driven scanning tunnelling spectroscopy on the ångström scale paves the way for ultrafast measurements of wavefunction dynamics in atomically precise nanostructures and future optoelectronic devices based on locally tailored electronic properties.

No abstract

Terahertz scanning tunneling microscopy (THz-STM) enables ultrafast measurements of surfaces, single molecules, and nanostructures with simultaneous sub-picosecond temporal resolution and atomic spatial resolution. In pump-probe THz-STM experiments employing femtosecond optical pump pulses, lightwave-driven tunneling by a time-delayed THz probe pulse accesses the evolving differential conductance of the tunnel junction following photoexcitation. However, a general theoretical approach to extract the time-and voltage-dependent differential conductance from THz-STM measurements is lacking. Here, we introduce an algorithm for pump-probe THz scanning tunneling spectroscopy (THz-STS) analysis. Our approach allows us to reliably reconstruct the tunnel junctions differential conductance in steady-state or time-dependent scenarios from simulated THz-STS data. The algorithm achieves subcycle time resolution, which we demonstrate by retrieving dynamics faster than the bandwidth of the input THz voltage transient. Subcycle THz-STS will make lightwave-driven microscopy yet more powerful as a tool for characterizing ångström-scale ultrafast dynamics in novel materials.

No abstract