Light-Field Photocurrent Spectroscopy

• Light-Field-Driven Photocurrent Spectroscopy is a suite of methods using controlled optical fields to generate electrical currents that reveal detailed quantum dynamics beyond simple absorption.
• It integrates local photoexcitation, carrier relaxation, and nonlocal current collection to expose band structures, excitonic resonances, and symmetry-dependent properties.
• Experimental approaches range from phase-tagged multidimensional coherent spectroscopy to plasmonic near-field techniques, showcasing advanced control over light–matter interactions.

Light-Field-Driven Photocurrent Spectroscopy denotes a broad class of spectroscopic methods in which the measured observable is an electrical photocurrent generated by a controlled optical or terahertz field, rather than solely an emitted optical field. Across this class, the driving field may be varied in photon energy, temporal modulation, pulse delay, phase, polarization, helicity, orbital angular momentum, spatial phase profile, or near-field distribution, while the resulting current is used to interrogate band structure, excitonic and polaritonic resonances, quantum geometry, carrier dynamics, local electrostatics, and nonlocal collection physics. In this sense, photocurrent is treated not merely as a device output but as a multi-physics diagnostic spanning photoexcitation, local current generation, relaxation, propagation, and collection (Ma et al., 2022).

1. Conceptual basis

Light-field-driven photocurrent spectroscopy rests on the fact that photocurrent generally encodes more than optical absorption alone. In the formulation emphasized for quantum materials, the measurable current integrates the full photoexcitation life-cycle: absorption, local current generation, relaxation and carrier dynamics, and propagation and collection. Because these stages occur on different spatiotemporal scales, photocurrent can be sensitive to optical selection rules, Bloch-wavefunction structure, hot-carrier kinetics, exciton formation, diffusion and drift, and device-scale collection geometry (Ma et al., 2022).

A central consequence is that photocurrent is often intrinsically nonlocal. The measured terminal current need not equal the local source current at the illuminated point, but can instead reflect the overlap between a local photocurrent density and a device weighting field. This is expressed by the Shockley–Ramo-type relation

$$I = A \int {\bf j}_{\rm loc}({\bf r}) \cdot \nabla \psi({\bf r}) \, d{\bf r},$$

which makes scanning and spatially structured experiments highly sensitive to device geometry and ambient conductivity as well as to the microscopic light–matter interaction (Ma et al., 2022).

This conceptual expansion is what distinguishes light-field-driven photocurrent spectroscopy from narrower photocurrent measurements that only track current versus wavelength. The family includes action-detected multidimensional coherent spectroscopy, gate-dependent terahertz photocurrent spectroscopy, intensity-modulated frequency-domain microscopy, cryogenic structured-light photocurrent spectrometry, and strong-field waveform-sensitive current readout, among others (Nardin et al., 2013).

2. Experimental architectures

The field comprises several distinct but technically related measurement architectures. In all of them, the optical field is shaped or scanned in a way that encodes specific dynamical or symmetry information into the photocurrent.

Before the table, two contrasts are useful. First, some architectures operate in the time domain with ultrafast pulse sequences and Fourier reconstruction, whereas others operate in the frequency domain through sinusoidal modulation or monochromatic THz excitation. Second, some methods exploit far-field illumination, while others use plasmonic or resonator near fields so that the local field texture, rather than the incident beam alone, determines the current.

Architecture	Controlled light field	Electrical observable
MD-COPS	Four collinear pulses $A,B,C,D$ with AOM radio-frequency tags and delays $\tau,T,t$	Complex fourth-order photocurrent $Z=X+iY$ and 2D rephasing/non-rephasing spectra
Pulse-shaper A-2DES	Phase-modulated four-pulse sequences with pattern parameters $N$, $n_i$, $N_{\mathrm{rep}}$	Fourier-isolated photocurrent at rephasing, non-rephasing, and double-quantum modulation frequencies
IMPS microscopy	Local AC intensity perturbation on top of DC light bias	Local complex transfer function $Q_{xy}(\omega)$
Gate-dependent THz PCS	Zero-bias CW THz illumination at multiple frequencies with gate-tuned $E_F$	Responsivity $R_I=I_{\rm pc}/P$ versus gate and $A,B,C,D$0
Cryogenic structured-light PCS	SAM- and OAM-controlled light from 500 to 700 nm under $A,B,C,D$1 T and 3 K	Spatially resolved photocurrent spectra versus helicity and topological charge
Strong-field waveform readout	Few-cycle or plasmonically enhanced optical fields with CEP or helicity control	Waveform-sensitive photocurrent and emitted-current asymmetry

In multidimensional coherent optical photocurrent spectroscopy, four collinear pulses generate a fourth-order population that is read out electrically as photocurrent rather than as a radiated phase-matched four-wave-mixing beam. Each pulse is tagged with a distinct AOM frequency, and the rephasing and non-rephasing pathways are selected electronically at

$A,B,C,D$2

with the complex signal recorded as $A,B,C,D$3 and Fourier transformed over $A,B,C,D$4 and $A,B,C,D$5 to yield multidimensional spectra. A copropagating cw reference provides passive phase-noise suppression, allowing unstabilized interferometers (Nardin et al., 2013).

Pulse-shaper-based photocurrent-detected 2D electronic spectroscopy implements a related four-pulse action-detected scheme but replaces external interferometers with an acousto-optic pulse shaper. There the central technical variables are the phase-stepping rule

$A,B,C,D$6

the pattern size $A,B,C,D$7, and the repetition count $A,B,C,D$8. The method isolates rephasing, non-rephasing, and double-quantum photocurrent components in Fourier space, but the same work also shows that improper trimming, incomplete discharge between pulse sequences, and pulse-shaper nonlinearities can distort the retrieved 2D spectra (Amarotti et al., 3 Jun 2025).

At the opposite end of the timescale spectrum, intensity-modulated photocurrent spectroscopy measures the complex transfer function

$A,B,C,D$9

with local imaging implemented by rastering a modulated beam over a device. In microscopy mode, the quantity becomes $\tau,T,t$0, so that frequency-domain transport information is mapped spatially rather than being averaged over the whole sample (Laird et al., 2022).

3. Field engineering and symmetry control

A defining characteristic of the field is that the driving light is often itself the experimental control parameter. The photocurrent is then interpreted as a response to field structure rather than to intensity alone.

For structured monochromatic fields in a two-dimensional electron gas, the dc current can be decomposed as

$\tau,T,t$1

where the three terms are driven respectively by intensity gradients, gradients of Stokes polarization parameters, and gradients of optical phase. In this description, radial and azimuthal photocurrents under twisted light become direct probes of local field intensity, polarization texture, and phase winding (Gunyaga et al., 2023).

Bicircular and bichromatic drives extend this symmetry engineering into coherent-control regimes. One theoretical route uses bicircular light made from counter-rotating $\tau,T,t$2 and $\tau,T,t$3 components, which induces dc photocurrent even in centrosymmetric systems through a dynamically generated effective polarity. In the weak-field regime, the resulting injection-type current scales as

$\tau,T,t$4

with explicit phase control, including $\tau,T,t$5 in the reduced one-dimensional result and $\tau,T,t$6 in a $\tau,T,t$7-symmetric two-dimensional case (Ikeda et al., 2023).

A related but more symmetry-selective development uses combinations of bichromatic linearly polarized $\tau,T,t$8–$\tau,T,t$9 beams. For special choices of relative polarization angle and two-color phase, the total tailored field preserves TRS while breaking the other crystal symmetries that would otherwise forbid photocurrent. In a TRS-invariant material, this imposes a forbidden photocurrent selection rule; in a TRS-broken phase, the node is lifted, yielding a background-free signal of intrinsic TRS breaking. The same work validates the mechanism with real-time \textit{ab initio} simulations and proposes its use for magnetism and Chern physics without external magnetic fields or circularly polarized probe fields (Lesko et al., 8 Jul 2025).

Cryogenic structured-light photocurrent spectrometry generalizes such control to both spin and orbital angular momentum. In monolayer $Z=X+iY$0-MoS$Z=X+iY$1, the instrument supports SAM $Z=X+iY$2 and OAM $Z=X+iY$3, under $Z=X+iY$4 T and temperatures down to 3 K, with about $Z=X+iY$5m spatial resolution. The reported measurements show monotonically increasing photocurrents with increasing $Z=X+iY$6, alongside excitonic Zeeman splitting and an enhanced Landé $Z=X+iY$7-factor attributed to enhanced formation of intervalley dark excitons under magnetic field (Hao et al., 30 May 2025).

Near-field engineering pushes the same logic into plasmonic and petahertz regimes. In one case, an inversely designed gold nanostructure uses local field amplitude, local helicity, ellipticity, and spatial gradients to generate a femtosecond in-plane drift photocurrent through an optical-rectification mechanism associated with the inverse Faraday effect (Mou et al., 2024). In another, monolithic arrays of plasmonic bow-tie nanoantennas convert few-cycle optical waveforms into CEP-sensitive photocurrent on an attosecond temporal scale, with the current oscillating at the carrier-envelope-offset frequency and scaling across electrically connected arrays (Yang et al., 2019).

4. Spectroscopic observables and information content

The central value of these methods lies in what the photocurrent can reveal. Depending on the field protocol, it can expose ultrafast coherence, local electric field, dynamic transport parameters, miniband topology, or cavity-mode structure.

In semiconductor nanostructures, photocurrent-detected multidimensional coherent spectroscopy recovers the complex nonlinear signal directly, including both amplitude and phase. In the demonstrated double InGaAs/GaAs quantum well, simultaneous rephasing and non-rephasing 2D spectra show a dominant diagonal excitonic peak whose slight elongation in the rephasing map is interpreted as evidence for slight inhomogeneous broadening (Nardin et al., 2013).

In GaN p–n diodes, near-edge photocurrent responsivity is used as a local electrostatic probe through the exciton Franz–Keldysh effect. After normalization, the spectrum is fit with an XFK-based model whose only adjustable parameter is the local bias $Z=X+iY$8, from which the local depletion widths and peak electric field follow uniquely. The extracted $Z=X+iY$9 varies linearly with applied bias, the spectroscopy-derived built-in bias is $N$0 V, independent C–V gives $N$1 V, and the overall accuracy of spectrally extracted $N$2 is estimated as about $N$3 V (Verma et al., 2020).

In photovoltaic and perovskite systems, frequency-domain photocurrent reveals transport and degradation physics inaccessible to static current mapping. IMPS microscopy measures local $N$4, enabling extraction of ambipolar diffusion-length maps and RC-limited high-frequency contrast. In the demonstrated degraded MAPbI$N$5 back-contact device, the inferred ambipolar diffusion length varied roughly from $N$6 to $N$7 nm, compared with pre-degradation values of approximately $N$8–$N$9m, while high-frequency contrast was interpreted primarily as a proxy for local inverse series resistance (Laird et al., 2022).

Gate-dependent THz photocurrent spectroscopy in nearly aligned graphene/hBN moiré superlattices turns the photocurrent map $n_i$0 into a miniband spectrometer. Off resonance, when $n_i$1 is smaller than the relevant gap, the response is intraband and enhanced near superlattice Dirac points because of lower Fermi velocities and specific valley degeneracies. Above gap, bulk photocurrent appears and is assigned to shift current associated with the geometric Berry phase of the minibands. Using frequencies from 0.075 to 4.7 THz, the method resolves avoided crossings and inversion-breaking local and global gaps in the $n_i$2 meV range that are described as inaccessible by conventional electrical or optical techniques (Delgado-Notario et al., 22 Jul 2025).

Photocurrent can also act as a local probe of cavity field structure. In a GaAs/AlGaAs 2DEG coupled to split-ring resonators, photocurrent spectroscopy resolves ultrastrong Landau-polariton coupling not only to bright SRR modes but also to optically dark dimer modes and topological edge modes in an SSH-like SRR chain. Reported normalized couplings include $n_i$3 for bright, dark, and edge modes, showing that the electrical readout tracks the local resonator near field rather than only the far-field oscillator strength (Huang et al., 27 Jan 2026).

5. Materials and representative regimes

The materials base of light-field-driven photocurrent spectroscopy is unusually broad. Semiconductor nanowires, quantum wells, contacted quantum dots, carbon nanotubes, and nanotubes appear as natural targets for electrically detected ultrafast spectroscopy because photocurrent readout remains practical where phase-matched radiated four-wave-mixing beams do not (Nardin et al., 2013). In the nanowire literature more broadly, photocurrent spectroscopy is used to probe morphology, band structure, interfaces, and ultrafast generation channels, with conventional and time-resolved methods separating photovoltaic, photothermoelectric, photogating, photodoping, photo-Dember, displacement-current, and persistent-photoconductivity contributions (Erhard et al., 2014).

Device-scale implementations include GaN diodes, molecular junctions, organic and hybrid-inorganic photovoltaics, back-contact perovskites, graphene photodetectors, and moiré superlattices. In molecular electronics, combined optical absorption and zero-bias photocurrent spectroscopy separate internal photoemission from molecular-absorption-mediated photocurrent and identify communication length scales of order $n_i$4 nm and $n_i$5 nm under different regimes (Mukundan et al., 2021). In graphene-on-gold gap devices, photon tunneling into surface plasmons enhances photocurrent, with resonant p-polarized excitation producing a significantly amplified response relative to s polarization and a photocurrent polarity reversal as the beam crosses the gap (Maleki et al., 2016).

Quantum-material implementations place particular emphasis on geometry, topology, and nonlocality. Reviews of the field highlight bulk photovoltaic effects, photothermoelectricity, ultrafast photocurrent autocorrelation, THz time-domain photocurrent spectroscopy, and scanning or near-field photocurrent nanoscopy as routes to probing Berry curvature, shift vector, quantum metric, topological invariants, and carrier-scattering pathways (Ma et al., 2022). In Weyl semimetals, simultaneous scanning photocurrent microscopy and NV-center quantum magnetometry show that anisotropic photothermoelectric current can dominate both bulk and edge photocurrent patterns, with long-range terminal currents arising through Shockley–Ramo collection rather than directly revealing the local source mechanism (Wang et al., 2022).

Timescales span many orders of magnitude. IMPS and CW THz measurements operate in the frequency domain from hundreds of hertz to kilohertz or terahertz; multidimensional coherent photocurrent spectroscopy resolves femtosecond coherent evolution; petahertz bow-tie arrays and plasmonic drift-current structures operate in the few-femtosecond or attosecond-response regime (Laird et al., 2022).

6. Interpretive issues, limitations, and methodological discipline

A recurring interpretive issue is the misconception that photocurrent directly measures local absorption. Several strands of the literature show otherwise. In quantum-material devices, the collected current is often a weighted nonlocal quantity rather than the local source. In anisotropic Weyl semimetals, scanning photocurrent signals that might be ascribed to bulk photovoltaic response can instead be explained by anisotropic photothermoelectric generation convolved with weighting-field collection (Wang et al., 2022). This suggests that photocurrent spectroscopy is often most reliable when local generation and global collection are analyzed together.

Another recurring issue is that electrical readout introduces device and circuit dynamics into the spectroscopy. Pulse-shaper A-2DES identifies phase leakage from incorrect trimming, signal accumulation from insufficient discharge between pulse sequences, startup irregularity when the laser repetition rate exceeds the pulse-shaper streaming rate, and distortions from elevated RF streaming power. The same work gives explicit mitigation strategies, including exact selection of integer modulation cycles for Fourier analysis and phase corrections of the form $n_i$6 and $n_i$7 (Amarotti et al., 3 Jun 2025).

Model validity is similarly critical. In GaN diodes, the data are described only by excitonic Franz–Keldysh theory; the non-excitonic Aspnes model fails and even yields an unphysical built-in voltage in the supplementary analysis (Verma et al., 2020). In MD-COPS, passive cw-referenced phase compensation is sufficient for unstabilized interferometers, but the compensation is explicitly partial, and the maximum usable delay is limited by the coherence length of the cw laser or by the travel range of the delay stages (Nardin et al., 2013). In IMPS microscopy, low frequencies are contaminated by ionic motion, while high frequencies can be dominated by RC attenuation rather than intrinsic recombination or diffusion (Laird et al., 2022).

Finally, some of the most forward-looking demonstrations remain theoretical or simulation based. The inversely designed plasmonic antenna producing up to 400 mA transient photocurrent and tunable THz emission is a combined electromagnetic inverse-design and thermal-verification study rather than an experimental spectroscopy paper (Mou et al., 2024). The bow-tie CEP detector is experimental, but it also documents laser-induced reshaping and signal degradation at high pulse energy, underscoring how strongly waveform-sensitive photocurrent depends on nanometric field concentration and structural stability (Yang et al., 2019).

Taken together, these constraints do not narrow the field so much as define its methodological character. Light-field-driven photocurrent spectroscopy is most effective when the optical drive, the microscopic current-generation mechanism, and the device-scale collection process are all treated as integral parts of the measurement. Under that condition, photocurrent becomes a versatile spectroscopic channel for coherent nonlinear response, local electrostatics, structured-light selection rules, miniband geometry, and ultrastrong light–matter coupling.