Light-driven nanoscale vectorial currents

Abstract

Controlled charge flows are fundamental to many areas of science and technology, serving as carriers of energy and information, as probes of material properties and dynamics¹ and as a means of revealing^2,3 or even inducing^4,5 broken symmetries. Emerging methods for light-based current control^5-16 offer particularly promising routes beyond the speed and adaptability limitations of conventional voltage-driven systems. However, optical generation and manipulation of currents at nanometre spatial scales remains a basic challenge and a crucial step towards scalable optoelectronic systems for microelectronics and information science. Here we introduce vectorial optoelectronic metasurfaces in which ultrafast light pulses induce local directional charge flows around symmetry-broken plasmonic nanostructures, with tunable responses and arbitrary patterning down to subdiffractive nanometre scales. Local symmetries and vectorial currents are revealed by polarization-dependent and wavelength-sensitive electrical readout and terahertz (THz) emission, whereas spatially tailored global currents are demonstrated in the direct generation of elusive broadband THz vector beams¹⁷. We show that, in graphene, a detailed interplay between electrodynamic, thermodynamic and hydrodynamic degrees of freedom gives rise to rapidly evolving nanoscale driving forces and charge flows under the extremely spatially and temporally localized excitation. These results set the stage for versatile patterning and optical control over nanoscale currents in materials diagnostics, THz spectroscopies, nanomagnetism and ultrafast information processing.

Fig. 1. Directional photocurrents in symmetry-broken optoelectronic metasurfaces.

a, Illustration of an optoelectronic metasurface consisting of symmetry-broken gold nanoantennas on graphene. Femtosecond laser illumination stimulates vectorial photocurrents and consequent emission of ultrafast THz pulses. b, Measured (solid line) and simulated (solid fill) transmission spectra for two nanoantenna designs with resonances at 800 nm (blue) and 1,550 nm (orange). c, Measured incident-wavelength-dependent THz field amplitude (solid lines with data markers) and d.c. photocurrent (dashed line), as well as simulated field intensity (solid fills). Top insets: scanning electron micrographs of the fabricated nanoantenna elements. Bottom insets: simulated plasmonic-field enhancements. Scale bars, 200 nm. d, Measured THz time-domain signals emitted from the resonantly excited 800-nm (left) and 1,550-nm (right) metasurfaces, compared with those from 1-mm-thick ZnTe. Curves are offset for clarity. Inset: amplitude spectra (plotted on a logarithmic scale) revealing bandwidth out to 3 THz, limited by phase matching in the 1-mm ZnTe detection crystal. e, Incident-fluence-dependent THz field amplitude and d.c. photocurrent readout, along with a linear fit to low fluence (solid fill). a.u., arbitrary units. Source Data

Fig. 2. Polarization-dependent local responses and omnidirectional control.

a,b, SEM images of uniformly oriented (a) and Kagome (b) metasurfaces with 800-nm resonances (overall metasurface size, 1 mm²). Scale bars, 500 nm. Insets: simulated resonant plasmonic-field enhancements for different incident linear polarization angles (black double arrows), with the calculated net current direction indicated (blue single arrows). c,d, Measured x (red) and y (blue) components of the radiated THz field (solid lines with data markers) and photocurrent (dashed lines) for the uniformly oriented (c) and Kagome (d) metasurfaces with respect to the incident linear polarization angle. Calculated linear responses (solid fills) are shown for comparison, with + and − signs indicating lobe polarity. For clarity, a small residual y component is not shown in c. e, The Kagome metasurface shows nearly constant photocurrent magnitude (purple) and continuously rotatable direction (grey), consistent with analytic predictions. Source Data

Fig. 3. Global vectorial currents and THz vector beams.

a, SEM image of the central region of a radial vector metasurface, with arrows illustrating the expected radial photocurrent on circularly polarized excitation. The entire metasurface (1 mm overall diameter) is excited. Scale bar, 1 µm. b, Far-field spatial map of the radial THz vector field, showing the total measured THz field magnitude (colour map and length of white arrows) and direction (direction of white arrows) at the pulse peak. c, Measured (top) versus ideal (bottom) Hermite–Gaussian modes for the x and y field components of the radial THz vector beam. Arrows in bottom plots indicate THz field polarity at peak time. d, Lineout from the radial E_x image, showing the transient THz field as a function of x position, along with example THz time-domain waveforms illustrating the polarity reversal across the beam. Inset: corresponding THz field amplitude spectrum in the frequency domain, plotted on a logarithmic scale. e–g, The same as in a–c but for the azimuthal vector metasurface and THz vector beam. Source Data

Fig. 4. Local photothermoelectric driving force and resulting nanoscale charge flow.

a, Measured resistivity of a metasurface device as a function of gate voltage. Insets: chemical potentials of the gold-pinned and bare graphene regions for three gate voltages as indicated in b. b, Measured gate-dependent current (dashed line) compared with calculated ΔS₀ (solid fill). Inset: illustration of the gated device configuration, with tip-localized heating and corresponding net carrier motion. c, Spatial distribution of S₀. d, Snapshot of the calculated T_e distribution at the time of the incident pulse peak (0 fs). e, Photothermoelectric x acceleration field (along the principal nanoantenna axis) at V_g = 0 V and 0 fs. f, Hydrodynamic velocity field for the hole fluid at V_g = 0 V and 0 fs. Source Data

Extended Data Fig. 1. Linearly polarized THz emission from directional photocurrent.

a, Measured (solid line) and approximated (solid fill) THz time-domain waveforms for a uniformly oriented metasurface under circularly polarized femtosecond laser excitation, assuming a Gaussian photocurrent pulse (dashed grey line; 600 fs full width at half maximum) with E_THz ∝ −dj/dt. The true current and THz field is expected to be faster than shown here, owing to bandwidth limitations of the photoconductive antenna detector (see Supplementary Note 1). Inset, spatially mapped THz field showing the beam with approximately uniform linear polarization. A small y-polarized contribution and corresponding roughly 10° angular deviation from x axis may be attributed in part to residual sample tilt and imperfect alignment of the THz imaging system. b, The x (red), y (blue) and total (purple) THz field amplitudes measured (solid lines with data markers) as a function of sample orientation under circularly polarized incident optical pulses, compared with the expected cos(θ) (red fill), sin(θ) (blue fill) and constant (purple fill) behaviours, respectively. The skew in the y dataset is attributed to imperfect circular polarization and residual misalignment in the parabolic mirrors. Source Data

Extended Data Fig. 2. Wavelength-sensitive directional photocurrent response.

a, Measured (solid lines) and simulated (solid fills) transmittance spectra for the perturbed Kagome lattice, with white light linearly polarized along the three nanoantennas and resonances nominally designed around 775 nm (0°; cyan), 825 nm (120°; orange) and 875 nm (240°; red). Inset, unit cell (scale bar, 250 nm). b, Direction of THz field polarization (and thus net photocurrent) measured (solid circles) and simulated (solid fills) as a function of incident wavelength for circularly polarized incident light. Bottom insets show simulated fields and net current directions for circularly polarized excitation at the three resonant wavelengths. Measured (c) and simulated (d) x (red), y (blue) and total (purple) THz fields as a function of incident linear polarization angle at the three resonance wavelengths, showing preferred alignment along the active resonance axes (dotted white lines). Source Data

Extended Data Fig. 3. Radial metasurface device and resonant excitation.

a, Optical micrograph of a radial metasurface, prepared as a device for photocurrent readout. Inset, radially outward-oriented nanoantennas with resonance at 800 nm and nanoantenna spacing about 500 nm. b, Measured DC photocurrent (dashed line) and simulated plasmonic hotspot field intensity (solid fill; same as in Fig. 1c), verifying the resonantly coupled/enhanced radial photocurrent. Source Data

Extended Data Fig. 4. Ultrafast time evolution of local thermodynamic and hydrodynamic quantities.

Calculated electron temperature, photothermoelectric acceleration field, hydrodynamic flow profile and relative scattering rates within the square metasurface unit cell as a function of time with respect to the 100-fs laser pulse. All plots are normalized to the peak values.

Extended Data Fig. 5. Electrostatically gated metasurface.

a, Metasurface device, with false colouring indicating the 3-nm/50-nm Ti/Au (light yellow), nanoantenna array (darker yellow), graphene (dark grey), 30-nm Pt back-gate electrode with 30-nm silica spacer layer (light grey) and substrate (light blue). b, Measured (solid line) and simulated (solid fill) reflectance spectrum of the device, showing resonant absorption around 875 nm. c, Gate-dependent photocurrent as a function of position, with locations indicated in a. Beam spot about 25 µm. To isolate the electrode contributions, the left and right electrode measurements are performed on an array-free device, with little laser polarization dependence observed. A polarity reversal is observed between the left and right electrodes. By contrast, strong polarization dependence is observed on the array at the centre of the sample. Both metasurface and array-free devices have 250-µm square graphene between the electrodes (metasurface device is the same as in Fig. 4). Photocurrents are normalized to the peak positive or negative values. Source Data