Light intensity-induced photocurrent switching effect

The study addresses how light intensity can control the polarity and magnitude of photocurrents in semiconductor-based systems. Motivated by applications in photovoltaics, photocatalysis, and optoelectronics, the authors investigate mechanisms of photocurrent generation and switching. Prior work has shown photo-electrochemical photocurrent switching (PEPS) by varying applied potential or excitation wavelength, but the role of light intensity is often overlooked. The research aims to demonstrate and explain light intensity-induced photocurrent switching (LIIPS) in a ZnO-based ternary hybrid and to argue for the broader significance of intensity-dependent effects in photochemical systems.

Previous studies established photocurrent polarity control via potential or wavelength in materials such as TiO2, CdS, and BiVO4, and through surface modification with organic acceptors (e.g., anthraquinones) and carbon nanostructures. Reports exist of anomalous decreases in photocatalytic activity under low light intensity for metal-doped wide bandgap semiconductors (Pt@TiO2, ZnO, GaN:ZnO) and plasmonic heterostructures, often attributed to electron trapping at co-catalysts and recombination center dynamics. Superlinear photocurrent dependence on intensity has been linked to competing recombination centers and reduced recombination rates at high flux. The authors’ prior work on PEPS and hybrid materials-based optoelectronic logic provides context for exploring emergent behavior in ternary systems with two modifiers that can alter energy alignment, trap-assisted transport, and interfacial kinetics.

Materials and hybrid preparation: Fullerenols (COH; C60(OH)30–36) synthesized from C60 via alkaline H2O2/TBAH treatment in toluene, isolated by methanol precipitation and drying. Ternary hybrid CA–COH@ZnO prepared by dissolving COH in water (sonication), adding chloranilic acid (CA; 3,6-dichloro-2,5-dihydroxy-1,4-benzoquinone), homogenizing, then adding ZnO powder and further homogenizing to saturate adsorption sites. Electrodes were fabricated by drop-casting hybrids on ITO@PET substrates; typical film thickness 7.3 ± 2.8 µm. Conditioning (light soaking): Prior to photoelectrochemical measurements, select samples underwent irradiation at 365 nm with an LED matrix at 125 mW cm−2 while held at a chosen potential (light-soaking pre-treatment). Conditioning potentials were varied within a window where LIIPS was observed; in most tests −300 mV vs Ag/AgCl was used for measurements. Photoelectrochemical measurements: Three-electrode setup with Ag/AgCl reference and Pt counter in 0.1 M KNO3 (pH ~6), electrolyte equilibrated with air or oxygen/argon as specified. Photocurrent action maps measured using a xenon lamp and monochromator; LIIPS probed using a 365 nm LED at controlled intensities. Photocurrent polarity versus light intensity and potentials was mapped (Fig. 2), including different conditioning potentials and measurement potentials. Spectroscopic and structural characterization: UV–Vis (solutions and diffuse reflectance of powders), ATR-FTIR (powders), Kelvin probe surface photovoltage spectra (SPV), SEM with EDS, and UPS to determine work function and band alignment. DPV used to detect new redox features after conditioning. Electrochemical analyses: Mott–Schottky analysis for flat-band potentials and charge carrier density. Light intensity-dependent electrochemical impedance spectroscopy (EIS) performed under 365 nm illumination (0.1, 0.5, 1.0 mW cm−2) at −300 mV vs Ag/AgCl to extract recombination resistance and chemical capacitance and compute electron lifetime τ = Rrec CCPE. Computations: DFT and TD-DFT (B3LYP/3-21G(6D,7F)) to model CA–COH interactions via hydrogen bonding vs covalent bonding and to simulate UV–Vis spectra and HOMO–LUMO charge-transfer transitions, informing interpretation of spectral changes upon irradiation and aging.

• Discovery of light intensity-induced photocurrent switching (LIIPS) in a ternary hybrid composed of ZnO, chloranilic acid (CA), and fullerenols (COH). At −300 mV vs Ag/AgCl in air-equilibrated electrolyte and 365 nm excitation, low intensities yield cathodic photocurrents while high intensities yield anodic photocurrents; a distinct intensity threshold separates the regimes.
• Necessary conditions for LIIPS: (i) both modifiers CA and COH must be present; (ii) a light-soaking (conditioning) pre-treatment under applied potential is required; (iii) dissolved oxygen must be present (air-equilibrated electrolyte). No LIIPS was observed without irradiation during conditioning or in oxygen-depleted conditions.
• Persistence and tunability: The LIIPS effect is persistent with respect to the direction of intensity change and is only partially affected by extensive cyclic voltammetry. The magnitude and completeness of switching depend on conditioning irradiation time; shorter soaking (<10 min) leads to a transitional regime with less pronounced switching.
• Spectroscopic and computational evidence of modifier interplay: FTIR shows shifts in CA C=O stretching (~1600 cm−1) upon mixing with COH, consistent with H-bonding. TD-DFT predicts a ~550 nm charge-transfer band for H-bonded CA–COH that disappears upon covalent bonding. Experimentally, fresh CA–COH mixtures exhibit a ~550 nm band that vanishes after irradiation or aging, indicating light-induced transformation to a covalently bound CA–COH conjugate. UV–Vis at low pH indicates enhanced COH agglomeration in presence of CA (~265 nm band).
• Electrochemistry: A new DPV peak appears near −400 mV vs Ag/AgCl after conditioning, consistent with formation of new redox-active states associated with CA–COH covalent linkage. Photocurrent maps show amplified cathodic currents and an anodic shift (~0.1 V) of the PEPS switching potential when both modifiers are present.
• Band alignment and energetics: For neat ZnO, flat-band potential is −0.92 V vs NHE; conduction band potential ~ −0.41 V vs NHE and valence band ~ +2.79 V vs NHE. UPS-derived work functions: ZnO 3.9 eV; CA@ZnO 3.0 eV; CA–COH@ZnO 3.1 eV. Ternary composite shows a 0.4 eV negative shift of Fermi level relative to HOMO (from 1.8 to 1.4 eV), consistent with reduced electron density and enhanced oxygen reduction at cathodic potentials.
• Light-intensity dependent EIS lifetimes (τ) at −300 mV vs Ag/AgCl (365 nm): • ZnO (not conditioned): τ = 1.17 s (0.1 mW cm−2), 0.65 s (0.5), 0.51 s (1.0) • ZnO (conditioned): τ = 1.08 s (0.1), 0.92 s (0.5), 0.76 s (1.0) • CA@ZnO (not conditioned): τ = 0.40 s (0.1), 0.51 s (0.5), 0.63 s (1.0) • CA@ZnO (conditioned): τ = 1.82 s (0.1), 1.66 s (0.5), 1.30 s (1.0) • CA–COH@ZnO (not conditioned): τ = 4.89 s (0.1), 6.33 s (0.5), 6.51 s (1.0) • CA–COH@ZnO (conditioned): τ = 1.97 s (0.1), 4.80 s (0.5), 6.61 s (1.0) These results indicate additional electron traps and intensity-dependent filling; after conditioning the ternary hybrid shows very long lifetimes at high intensity, correlating with anodic response as traps saturate.
• Mechanistic model: At low light flux, COH-related traps (enabled by covalent CA–COH linkage) are largely empty and efficiently mediate interfacial electron transfer to O2, favoring cathodic photocurrents (e− + O2 → O2−, E° ≈ +0.33 V vs NHE at pH 6). At high flux, limited trap density leads to saturation; electrons are transferred to the substrate and water/hydroxyl oxidation dominates (H+ + O− → OH−, E° ≈ +2.03 V vs NHE at pH 6), yielding anodic photocurrents. A saturation-type kinetic expression i = i0(1 − e−kΦ) captures intensity dependence and the crossover between competing pathways.
• Morphology: SEM/EDS show no significant morphology changes due to modifiers or conditioning; slight increase in agglomeration tendency in ternary hybrid.
• SPV and PEPS: All samples exhibit n-type SPV signatures; ternary hybrid shows distinct SPV shape and relaxation behavior, consistent with altered carrier accumulation and trap-mediated dynamics. CA introduces PEPS (potential/wavelength switching) with amplified cathodic currents in oxygen-rich conditions; COH further enhances cathodic response and shifts switching potential slightly anodically.

The findings demonstrate that light intensity, not only potential or wavelength, can dictate photocurrent polarity in semiconductor hybrids. The LIIPS effect arises from a synergy between a semiconductor charge generator (ZnO) and dual modifiers (CA and COH) that withdraw and trap electrons. Light-induced chemical transformation converts H-bonded CA–COH into a covalently linked conjugate anchored to ZnO, creating an efficient electron-transfer pathway to oxygen at low photon flux. As intensity increases, finite trap states saturate, decreasing the relative rate of oxygen reduction and enabling anodic processes (water/hydroxyl oxidation) to prevail, reversing photocurrent polarity. Electron lifetime trends corroborate slow recombination and trap-controlled discharge dominated by interfacial electron transfer. This intensity-governed competition between two one-electron interfacial processes explains the observed maps of photocurrent vs potential and light flux and aligns with observed DPV and UPS shifts. The work underscores that intensity effects and trap-state dynamics are critical in photoelectrochemical systems and may broadly influence photocatalytic efficiency and device behavior.

This work introduces light intensity-induced photocurrent switching (LIIPS) in a ZnO-based ternary hybrid modified with chloranilic acid and fullerenols. The effect requires both modifiers, oxygen, and a light-soaking pre-treatment that induces covalent CA–COH linkage, establishing trap-mediated electron pathways. A mechanistic framework based on limited-density traps and saturation kinetics explains polarity reversal: cathodic oxygen reduction dominates at low flux, while anodic oxidation takes over at high flux. The study highlights underestimated roles of light intensity and trap-state kinetics in photocatalysis and optoelectronics. Future research should quantify trap densities and energetics, explore different modifier chemistries and loadings to tune LIIPS thresholds, extend to other semiconductors and co-catalyst systems, and integrate operando spectroelectrochemistry to directly observe interfacial intermediates under varying light intensities.

• The system is complex (ternary composite with light-induced chemical transformation), making a fully self-consistent mechanistic assignment challenging. Direct identification of the covalent linkage and intermediate species is inferred from spectroscopy and computations rather than unambiguous structural determination.
• LIIPS requires pre-irradiation conditioning and oxygen; generality to oxygen-free processes or other redox environments is not demonstrated.
• No photosensitization in action spectra despite added absorptions suggests unresolved aspects of light–matter interactions in the hybrids.
• The effect could not be fully reversed by electrochemical cycling within the accessible window, limiting dynamic reconfigurability.
• Quantitative surface coverage and exact trap state densities are not determined due to complex interactions in the hybrid.